Professional Documents
Culture Documents
JEFFREY PFEFFER
9 8 7 6 5 4 3 2 1
Printed in the United States of America
on acid-free paper
PREFACE
I would probably have never written this book or the Handbook of Social Psy-
chology chapter that was its progenitor had the decision been a rational one.
Since the last time I had attempted such an effort in the early 1980s, the field
of organizational behavior had continued along the path of increasing growth,
differentiation, and paradigm proliferation that made undertaking any sort of over-
view or review of the field an "almost impossible" effort—a phrase used repeat-
edly by colleagues who were kind enough to provide advice and comments on
the attempt. But within days of receiving the invitation to do the chapter for the
fourth edition of the Handbook, I learned I had almost a complete blockage of
the coronary arteries feeding the left side of the heart (which meant a "coronary
event" —I love the language of physicians —would be instantly fatal) and, conse-
quently, I needed open heart surgery. "What should I do about this invitation?" I
asked my wife and best friend, Kathleen. "Say yes," she replied. "That will give
you something to look forward to and a feeling that you will have a future." It
seemed like a good idea at the time, so 1 did say yes, and following all of the
principles of escalating commitment that we know so well, wound up writing first
a long chapter and then this book.
The task proved to be a formidable one, and I am quite resolved not to do it
again. Making sense of such a diverse field and deciding what to cover and what
to leave out is invariably both difficult and an undertaking that can never fully
please anyone except the author, and not even that person on a regular basis.
When I sent out the book manuscript for comments, and when I sent out a
working-paper version of the chapter (completed before the book, although who
knows if it will have appeared in print before the book does), I invariably got
what Bob Sutton came to call the "me, me, me" response. "This is a Rorschach
test," said Sutton, and he was, as usual, right. The modal comment I received
was something along the lines of "why didn't you include more [or anything]
vi Preface
about my favorite topic, which is also among the most important in the field?"
And, "how could you have overlooked citing some of my most important and
very best papers?" which the authors of these messages were often kind enough
to enclose so I wouldn't have to expend too much effort tracking them down. At
first I took all of this as a compliment—surely people must think this writing will
be influential and important to be willing to spend effort lobbying for the inclu-
sion of their favorite material. After a while, however, I realized it reflected an
egocentric view of the field that is an almost inevitable outcome of the absence
of agrecd-upon paradigms and frameworks. Years ago Jerry Salancik and I had
shown that when there was uncertainty to resolve, social similarity came to play
an important role in decision processes. Given the prevalent uncertainty and ab-
sence of agreement about the content and direction of organization studies, social
similarity would again be expected to play a role in people's views of the field,
and no one is more similar than ourselves.
So, let me begin with an apology. If I have failed to include your favorite
work or to cite you sufficiently, it is not because I am necessarily ignorant of that
part of the field or of your work, although I might be. It is because to keep this
effort to manageable proportions, decisions did have to be made about what to
leave in and what to leave out. These decisions were premised on my goal to
provide an overview of the field for new entrants such as graduate students and
scholars from adjacent disciplines. That implied focusing on what constitutes the
most basic and fundamental questions about how to understand organizations
and directing attention to what are some (not all) of the topic areas that have
drawn increasing research attention and controversy. I make no claims for the
ultimate wisdom of these choices—I did the best I could. I invite readers who
don't like the choices I made to supplement this material with their own.
I very much appreciate the helpful comments on drafts of this material pro-
vided by Pain Haunschild, Joanne Martin, Charles O'Reilly, Bob Sutton, Michael
Tushman, and Karl Weick. I have listened to their comments even if I have not,
in every instance, followed their advice. I also appreciate the comments of Marjo-
rie Williams of Harvard Business School Press, who was kind enough to read the
manuscript even though the press did not wind up publishing it. The editors of
the Handbook of Social Psychology, particularly Dan Gilbert, were very helpful
in their guidance. And, of course, thanks to Herb Addison, my editor at Oxford
University Press, for his guidance and confidence in the project. It has been a joy
to work with Herb and the press, and I hope they feel the same way.
My life is made easier by my incredibly wonderful assistant, Katrina Jaggears,
who did whatever I asked to help with the manuscript. My existence is made
joyous and wiser by my beloved, my bride, my wonderfulness, the Amazing Kath-
leen. And, in a sense more real than I sometimes like to think about, my complet-
ing this book (and everything else since the fall of 1993) was made possible by
Drs. James Avery and Joel Klompus, each of whom has both exceptional technical
skill and a dedicated, total quality approach to medicine. All things considered, I
am lucky indeed.
CONTENTS
References, 205
Index, 245
New Directions for Organization Theory
This page intentionally left blank
1
The Development and Scope
of Organization Studies
3
4 New Directions for Organization Theory
revolution has received little study . . . It has crept upon us silently. (1968, pp.
3-4)
And in times past, fewer people worked for others: "We look back after wage
labour has won a respected position by two centuries of struggle. We forget the
time when complete dependence on wages had for centuries been rejected by all
who regarded themselves as free men" (Hill, 1964, p. 63).
One of the indicators of the importance of organizations is that with increas-
ing frequency "organizations are singled out as the source of many of the ills
besetting contemporary society" (Scott, 1992, p. 4). For good or ill, many of soci-
ety's assets are managed by and controlled by organizations such as mutual funds,
banks, and pension funds. Their decisions affect who will get credit, what invest-
ments and technologies will receive backing, and, as a consequence, which see-
tors of the economy will be developed, and how and where people live. "Orga-
nized" religion has spread, as have organi/ed community and self-help groups, to
do things that used to be done informally by neighbors in a community. Through
their actions, these organizations define poverty and social problems as well as
help determine their remediation. The fact that so much of our material and
social welfare and life is inextricably bound up with organizations means that it
is important to understand how they function and how they can be analyzed. As
Stern and Barley have argued, it is important to study "how organizations affect
the social systems in which they are embedded" (1996, p. 146).
The field of organization studies, developed to understand these ubiquitous
social entities, comprises an interdisciplinary focus on (a) the effect of social orga-
nizations on the behavior and attitudes of individuals within them, (b) the effects
of individual characteristics and actions on organizations, with a particular em-
phasis on the efficacy and, indeed, the possibility of potent individual influence
(e.g., through leadership) in organizational systems, (c) the performance, success,
and survival of organizations, (d) the mutual effects of environments, including
resource and task, political, and cultural environments on organizations and vice
versa, and (e) concerns with both the epistcmology and methodology that under-
gird research on each of these topics. As such, the study of organizations is broad
in both its theoretical scope and empirical focus. For instance, social psychologi-
cal theories and findings relevant to the above issues would appear as a subfield,
and possibly a small one at that, in the domain of organization studies.
Given the breadth and scope of the field, which today reaches out to the
humanities (Zald, 1993), one might well ask to what extent considering organiza-
tion studies as a separate subject of inquiry makes sense, especially since the
discipline's boundaries are so permeable. This question has traditionally been an-
swered by the claim that the interdisciplinary nature of organization studies makes
cross-level analysis and the advance of theory more likely through processes of
cross-fertilization. However, there are clear trade-offs involved. Even as organiza-
tion studies becomes broad enough in scope to include virtually all relevant
methods and theoretical perspectives, this very breadth makes integration of
knowledge or even cognizance of the entire domain of organization studies sub-
stantially more difficult.
Development and Scope of Organization Studies 5
Because of the scope and diffuseness of the subject matter, which has in-
creased substantially over time, any treatment of organization studies must neces-
sarily be somewhat arbitrary in its coverage and its organization. In previous incar-
nations of this book (Pfeffer, 1982; 1985), I chose to organize the material by-
focusing on different levels of analysis (individuals, coalitions, and subunits on
one hand and the total organization on the other) and on different perspectives
on action. This latter categorization differentiated among perspectives that em-
phasized rational choice and decision making, situational constraint and social
influence, and a third perspective focusing on the emergent, almost random na-
ture of organizational behavior. Although treatments by level of analysis are still
common, particularly in texts on organizations (e.g., Northcraft and Neale, 1994;
Wagner and Hollenbcck, 1992), and although differences in perspectives on ac-
tion remain important, for the researcher or student trying to access the field
what may be even more important are the fundamental concepts and models
used and useful for understanding organizations and the controversies that charac-
terize the development of the field. Consequently, what I have chosen to do in
this book is first to give the reader some sense of the evolution and particularly
the more recent history of the field and its scope and content and then to con-
sider the relevant literature organized by the major issues and concepts that make
sense of organizations and are relevant to those who live in and manage them.
r
lb understand or analyze organizations, it is important to consider the locus
of causality and whether that is lodged in individuals, situations, or some combi-
nation. The locus of causality directs where we place our research emphasis and
how we go about understanding and affecting behavior. This topic constitutes the
focus of chapter 2. It is the case that scholars, managers, and casual observers
approach the analysis of organizations with a set of implicit or explicit models of
behavior. These models both structure what we observe and learn and also affect
the choice of how to intervene to change organizations and their members. Chap-
ter 3 explores five various models and assumptions about behavior that distinguish
some prominent, competing theories and approaches to organizational analysis,
including the economic model, the social model, the retrospectively rational
model, the moral model, and a cognitive, interpretive model of organizational
behavior.
As developed in chapters 2 and 3, one important distinction among theories
of organizations is the extent to which dimensions of the social structure are used
to help understand behavior by analyzing the context and constraints within
which behavior occurs. Chapter 4 presents an overview of the literature on demo-
graphic composition and its effects on organizations, one prominent example of
the use of structural analysis to understand organizations and groups within them.
Chapter 5 then considers mechanisms of social control. Control has been a major
theme in organizational studies since it began more than 100 years ago and con-
tinues to be prominent in numerous literatures relevant to organizations, includ-
ing culture and socialization, rewards and incentives, and leadership. Control has
a hierarchical emphasis to it, but many of the decisions in organizations occur
through processes of interpersonal influence that rely on power and negotiation
rather than on formal systems and authority. Chapter 6 explores how power and
6 New Directions for Organization Theory
social influence are developed and exercised and how negotiation, one important
way in which interpersonal influence is employed, occurs. Since the beginning
of organizational studies, one of its goals has been to understand variations and
determinants of organizational performance. Chapter 7 considers some prominent
approaches to analyzing organizational performance and how they have evolved.
Subjects such as control and organizational performance speak to the manageri-
alist orientation that has been and still is prominent in organization studies. This
functionalist orientation has stimulated a reaction in critical theory, which seeks
to remind us that "organizations can trample personal freedom and individual
fulfillment" (Stern and Barley, 1996, p. 148). Chapter 8 provides a brief overview
of what constitutes some elements of a critical perspective on organizations and
organization studies. In the concluding chapter 9, some substantive topics, largely
overlooked in recent writing, are considered, and I take up the question of the
evolution of organization studies and whether or not we are increasing our under-
standing of these important social entities. Together these topics and their explora-
tion permit us to begin to understand what we know and what we don't, to iden-
tify some important directions for future research, and to glimpse the extent to
which research and writing has been unevenly distributed and some of the factors
associated with the field's attention.
A number of conclusions or themes emerge from this overview of the field.
First, the location of much of organization studies in business schools, and U.S.
business schools at that, coupled with the profusion of theories and absence of
paradigmatic consensus has led to a substantial overemphasis of economic models
and logic. The field is unduly entranced by economistic thinking. This is unfortu-
nate because social and social psychological modes of organizational analysis are
potentially both more valid and useful. Second, the location of organization stud-
ies has also led to a focus on particular subjects (e.g., performance, survival, lead-
ership, culture) and an emphasis on a more micro (individual or individual
organization) level of analysis. The relationship between organizations and society
and various social problems and issues is given only occasional attention (e.g.,
Stern and Barley, 1996). Critical perspectives are seldom seen in empirical analy-
ses and are more frequently found in European writings. But as organization
studies evolves in Europe and tends to be located increasingly in prominent
schools of administration, one wonders about the future of the European critical
tradition.
Third, the absence of paradigmatic consensus and the consequent value
placed on uniqueness (Mone and McKinley, 1993), cleverness, and being inter-
esting have not always led to research focusing on powerful but apparently obvi-
ous effects on organizational behavior. And, coupled with the occasional insecu-
rity that comes from paradigmatic profusion and the vulnerability that creates,
organization studies has often eschewed an engineering orientation, implied by
being an applied social science, that was present at its founding (e.g., Thompson,
1956). Tims, the field is at once captivated by topics that have the appearance of
being applied without always taking seriously issues of design, intervention, and
change.
Development and Scope of Organization Studies j
Organizations Defined
How are organizations distinguished from other social collectivities such as small
groups, families, mobs, and so forth? Scott noted that "most analysts have con-
ceived of organizations as social structures created by individuals to support the
collaborative pursuit of specified goals" (1992, p. 10). Parsons (1956) distin-
guished organizations from other social collectivities by noting that organizations
had some purpose or goal. Donaldson maintained that "organizations are created
and sustained . . . in order to attain certain objectives" (1995, p. 135). The goal-
oriented or instrumental view of organizations implies that organizations are col-
lections of individual efforts that are coordinated to achieve things that could not
be achieved through individual action alone (Pfeffer and Salancik, 1978, p. 23).
However, defining organizations in terms of goal pursuit is somewhat prob-
lematic, and there is a large literature that treats the concept of an organizational
goal and whether or not the construct is meaningful (e.g., Simon, 1964). One
problem is that many organizations have within them members or employees
who either do not know the organization's goal or, if they do know it, do not
necessarily support it. For instance, although the goal of most publicly held cor-
porations in the United States is something like profit maximization or the max-
imization of shareholder value, many employees are more concerned with job
security and their relative influence than they are about profits. In one contract
manufacturer, employees, not knowing the elements of corporate accounting,
thought that the more overtime they worked (and the higher their own incomes),
the better. The goal of maximizing overtime, however, was not an organizationally
sanctioned or even recognized goal. Moreover, the goal of maximizing share-
holder value does not generate much commitment or excitement among most
organizational participants.
A second issue is that there is evidence that even when a goal is clearly
identified, if and when it is attained organizations often develop new goals—as if
the goal of an organization, once created, was simply its own continued survival
and perpetuation. In a notable example of this process, Sills (1957) studied what
happened to the National Foundation for Infantile Paralysis (the March of
Dimes) when a vaccine for polio—the reason for its existence — was discovered.
As we know, the March of Dimes did not disband upon achieving this success
but instead took on the task of understanding and assisting in the treatment of
other diseases such as genetic birth defects that crippled children.
Or consider the case of the Interstate Commerce Commission (ICC), estab-
lished by legislation in 1887 to "break up huge concentrations of wealth in the
hands of the country's railroad magnates" (Sanger, 1996, p. 9). Although railroads
came to face substantial competition from trucking and the power of the railroad
magnates declined substantially, the organization's purpose evolved and it en-
dured. Over the years its powers expanded substantially so that it "gained author-
ity over trucks, buses, and virtually anything else that moved across state lines.
. . . consumer complaints about moving companies . . . had to go through the
I.C.C." (Sanger, 1996, p. 9). Although the deregulation movement of the late
1970s spelled its apparent demise, the ICC survived for some fifteen years after
8 New Directions for Organization Theory
virtually all of its functions were removed. Stripped of its power to set transporta-
tion rates or to control entry to the industry except in very rare instances, the
agency continued to collect reams of data on transportation rates and employ a
considerable staff:
connection to the state, which confers legitimacy that distinguishes them from
less formal collectivities such as families, mobs, or informal groups.
Organizations can also be distinguished by the nature of their boundaries.
Inclusion in an organization is something granted by that organization, frequently
with some sort of formal designation such as a membership card or employee
identification. One is born into a family, and expulsion from the family is unlikely
if not impossible. The boundaries of small groups and mobs arc clearly evanes-
cent. It is seldom the case in less formalized groups that the task of boundary
maintenance and demarcation becomes a significant and identified role, but this
is common in organizations— hence the frequent description of organizations as
being "formal." Although organizational boundaries are clearly permeable—after
all, an open-systems view of organizations sees them as taking in various types of
inputs and, after some transformation process, exchanging what is produced for
resources to continue the cycle (Katz and Kahn, 1978) — permeability is to some
degree under the control of the organization. Formal boundary-spanning units
such as purchasing and human resources attempt to ensure that inputs meet
specifications, while other boundary-spanning units such as marketing and public
relations seek to develop external support and demand for what the organization
does.
So, we can say that organizations are more likely than other social groups to
have a goal of survival and self-perpetuation, have more clearly defined, demar-
cated, and defended boundaries, and often (although not invariably) have some
formal relationship with the state that recognizes their existence as distinct social
entities —obligated to pay taxes as an entity, capable of suing and being sued, and
so forth.
Several trends are apparent in organization studies as the field has evolved since
the mid-1980s. First, the field is increasingly, although certainly not exclusively,
lodged in business schools. It has virtually disappeared in psychology and political
science departments, particularly in the United States, and, although it seems to
be enjoying some slight resurgence in sociology departments, still represents a
comparatively small part of that field also. The study of management and organi-
zations is currently growing, albeit from an extremely small base, in schools of
engineering, particularly in departments of industrial engineering and engi-
neering management. This concentration of organization studies in a professional
school environment is a phenomenon with important consequences for both
what is studied and how it is studied. And, the evolution of organization studies
in various disciplines is informative about the development of the field and its
orientations, for this history still affects many aspects of the content of organiza-
tion studies.
The study of management and organizations had some substantial part of its
origins in engineering, both mechanical and industrial. Shenhav documented the
move to standardize mechanical parts and processes and to professionalize the
io New Directions for Organization Theory
discipline of engineering in the late 1800s. He noted that "beginning in the late
1880s, parallel to the attempts to standardize and systematize mechanical matters,
the movement was extended more explicitly to organizational and administrative
issues" (1995, p. 560).
The rise of Frederick W. Taylor and scientific management in the early 1900s
marked the culmination of the efforts to apply engineering principles to the de-
sign and management of work and ushered in the golden age of engineering.
Between 1880 and 1920, the number of engineers in the United States grew from
7,000 to 135,000 (Jacoby, 1985).
and organizations "around engineering ideals rather than around religious, phil-
anthropic, paternalistic, or social Darwinist ones" (p. 564) also continues to the
present.
Although Taylorism and engineering remain influential in the literature, this
orientation is now more likely to be found in business schools. Management
concerns have diminished in schools of engineering. Industrial engineering de-
partments have substantially diminished in importance over time compared to
disciplines such as electrical, chemical, and civil engineering, which are based
on physical, not social, sciences. To the extent industrial engineering has survived,
much of its attention has shifted to production and operations management,
which it views from a modeling or mathematical perspective emphasizing tech-
niques such as queuing theory and linear programming. With only a few excep-
tions, the study of organizations and management diminished substantially in in-
dustrial engineering departments, which were themselves in some decline.
Since the 1980s, there has been some reversal in this trend, for two reasons.
First, business education, particularly in private schools, is often offered only to
graduate students. This permitted industrial engineering to offer engineering
management to undergraduates, an attractive option in the educational market
place. Second, concerns with the management of high-technology companies and
the frequent transition of engineers from technical to managerial responsibilities
as their careers progressed have led to a market-based demand for more emphasis
on management topics in engineering schools. Nevertheless, concern with man-
agement issues remains comparatively small in engineering colleges today, partic-
ularly when one contrasts the present with the situation around the turn of the
century, when engineering was taking a leading role in developing the study of
management and organizations.
The study of organizations began in psychology with industrial and organiza-
tional psychology, subspecialties within psychology departments. The Journal of
Applied Psychology, first published in 1917, contained articles on subjects in ap-
plied psychology—testing and measuring individual capabilities, interests, and at-
titudes through various assessment instruments (the study of individual differ-
ences), the effects of work environments on attitudes and, to a lesser extent,
behavior such as turnover and performance, and the design (both physical and
social) of jobs and work environments. Table 1-1 presents the titles of selected
articles from the first several volumes of the Journal of Applied Psychology to
illustrate both the enormous continuity in subject matter between then and now
and the substantive focus of the early research in applied psychology.
Braverman argued that both industrial psychology and industrial sociology
arose as handmaidens to the engineering orientation of Taylorism and can be
best understood in that context:
Shortly after Taylor, industrial psychology and industrial physiology came into
existence to perfect methods of selection, training, and motivation of workers,
and these were soon broadened into an attempted industrial sociology, the study
of the workplace as a social system. The cardinal feature of these various schools
12 New Directions for Organization Theory
. . . is that . . . they do not by and large concern themselves with the organiza-
tion of work, but rather with the conditions under which the worker may best be
brought to cooperate in the scheme of work organized by the industrial engineer.
(1974, p. 140)
Industrial and organizational psychology first lost status and then position
within the discipline, quite possibly because of its applied orientation and its role
in supporting management and engineering. 'Ibday industrial psychology exists in
relatively few psychology departments and virtually none of the most prestigious,
'['he Society for Industrial and Organizational Psychology (SIOP) has split off
from the American Psychological Association and now holds separate meetings.
And social psychology's increasing emphasis on individual cognition on the one
hand and personality on the other, with a decmphasis on groups and social influ-
ence (e.g., Markus and Zajonc, 1985) has left a growing gulf between more main-
stream psychological research and organizational issues and problems.
In sociology, the study of organizations was at one time a central focus of the
discipline. Many of the central figures in sociological theory—Weber (1947),
Marx (1967), and Durkheim (1949) —contributed to the understanding of bureau-
cracy, the employment relationship, and the division of labor and basis of solidar-
ity. A series of case studies of the Tennessee Valley Authority (Selznick, 1949), a
gypsum mine (Gouldncr, 1954), and a state employment agency and federal law-
enforcement organization (P. M. Blau, 1955) began the process of developing
empirically based general principles to describe organizational functioning. Al-
though organizational sociology remains viable, particularly in comparison to in-
dustrial/organizational psychology, from 1990 to 1994 only 15 percent of the arti-
cles in the American Sociological Review, the discipline's leading journal, could
be considered to concern organizations. Thus, organization studies' place within
sociology is a small one, particularly compared to the study of stratification.
A similar erosion of interest in organizations is visible in political science.
Many of the most prominent early figures in organization science came from
Table 1-1 Titles of Selected Articles from the 1917, 1918, and 1919 Volumes of the
Journal of Applied Psychology
political science —people such as Herbert Simon (1947) and James March
(March and Simon, 1958; Cyert and March, 1963). Many of the earliest studies
of organizations were studies of public bureaucracies (e.g., P. M. Blau, 1955; Blau
and Scott, 1962; Selznick, 1949), so there were many connections between public
administration, at one point an important subfield of political science, and organi-
zational behavior. However, as documented by Green and Shapiro (1994), the
discipline of political science has evolved to a point at which it has lost its institu-
tional roots (March and Olsen, 1989) and instead has come to be dominated by
formal models of rational choice. By 1992, this particular theoretical perspective,
rational actor theory, was represented in some 40 percent of the articles published
in the American Political Science Review, the disciplinary association's major jour-
nal. This formal modeling, often based on economic assumptions and method-
ological apparatus, has left political science increasingly distant from the less
mathematical methods that characterize theory and research in organization stud-
ies and also somewhat distant from the concerns of administration and practice.
Public administration, as a major subfield of political science, has withered as the
study of institutions has become less important in the discipline, replaced by an
emphasis on voting, coalitions, and legislative behavior.
The study of organizations fit uneasily in basic disciplines such as sociology
and psychology where it appeared to be too applied and became marginalized in
engineering schools, which were more concerned with disciplines based on the
physical sciences. But the study of management, administration, and leadership
has always found a home in business schools because of continuing interest in
subjects such as motivation, human effects on productivity and performance, and
organizational structure and strategy. Although organization studies programs have
at times been under attack in various business schools (e.g., Carnegie Mellon,
Yale, Rochester) because of the field's absence of mathematical, formal rigor,
demands for relevance from business schools' constituencies have ensured it a
place in business school curricula where it is a required course at both the under-
graduate and graduate levels.
Business schools are much more prevalent in the United States than in Eu-
rope, and as a consequence the location of organization studies primarily in busi-
ness schools has given it a distinctly American flavor. Clegg has noted that "orga-
nization studies . . . have drawn on both a range of materials and theoretical
approaches which have been too restricted" (1990, p. 1). His descriptions of the
French bread industry, Benetton, and industrial organization in the Far East in-
cluding Taiwan, Korea, and Japan consistently challenge ideas of efficiency,
whether these are derived from the transaction cost theory of Williamson (1975)
or structural contingency theory ideas (Donaldson, 1985). Clegg noted that "the
major thrust of efficiency arguments has been towards predicting a convergence
in the range of organizational forms to be found in modernity" (p. 151). The
proliferation of forms actually observed—witness the difference between the bread
industries of the United States and France—belies simple efficiency accounts.
But, these differences are apparent only to the extent one's empirical and theoreti-
cal scope transcends the confines of single countries, particularly the United
States.
14 New Directions for Organization Theory
Thus, the history and location of organization studies are consequential be-
cause the environment of behavior, including scholarship, has important effects
on that behavior. The increasing concentration of organization studies in business
schools or schools of management has had a number of important effects on the
development of the discipline. One such effect is the influence of adjacent disci-
plines. By being in a business school, organization studies has come to be located
in close institutional proximity to economics. Economics has a well-developed
scientific paradigm, considering it is a social science, with greater consensus con-
cerning what are the important issues to be studied and what are appropriate
styles of theory and methodological approaches (e.g., Pfeffcr and Moore, 1980a,b;
Lodahl and Gordon, 1972). Organization studies has no such paradigmatic con-
sensus (Pfeffer, 1993). Economics also has prestige, in part because of its extensive
use of mathematical formalism, and in part because many business schools (e.g.,
the University of California at Berkeley) emerged out of economics departments,
making that the mother discipline. As a consequence, economic ideas have come
to have significant influence on organization studies. The economic theories that
have been imported most frequently into organization theory—agency theory, hu-
man capital theory, and transaction cost economics —almost invariably proceed
from a theoretical position of methodological individualism, calculative rational-
ity, and presumptions of self-interested behavior and effort aversion (shirking).
As an example of economic theory's growth in prominence in the study of
organizations, consider the increase in the proportion of articles in two major
journals that cite economics, either transaction cost economics or some version
of agency theory. In 1975, just 2.5 percent of the articles in Administrative Science
Quarterly, the major journal in the field, and 0 percent of the articles in the
Academy of Management Journal, the publication of the major disciplinary associ-
ation, had citations to economic literature. Within ten years (1985), 30 percent
of the articles published in ASQ and 10 percent of the articles appearing in AM/
cited economics. By 1994, the percentages had increased to 45 percent and 28
percent, respectively, numbers strikingly similar to those tracing the growth of
rational actor models in political science. It seems almost impossible to consider
the evolution of organization theory, particularly in the future, without consider-
ing the influence of economics on the field's methods and substance.
The second effect of having organizations studies sited where professional
education (of MBA students) and practical concerns loom larger is that such
realities influence, even if only subtly, the topics chosen for study and the theoret-
ical lens brought to bear on them. For instance, although there have been a
number of critiques of the leadership literature (e.g., I.ieberson and O'Connor,
Development and Scope of Organization Studies 15
1972; Pfcffcr, 1977; Meincll, tthrlich, and Dukerich, 1985), some of which partic-
ularly question whether leadership has an effect or whether we just like to believe
in the controllability of the environment and hence have a "romance" with the
concept of leadership, research and writing on leadership continues apace. The
recent emphasis on charismatic leadership (House, 1977), the traits or personali-
ties of effective leaders (House, Spangler, and Wyocke, 1991; Winter, 1987), and
transformational leadership (Bass, 1985) — or leadership capable of accomplishing
large-scale, systemic change in organizations — reflects the effect of an environ-
ment in which leadership training is big business. For example, the Center for
Creative Leadership, which does leadership training as well as research and mate-
rials development on this subject, saw its revenues grow from about $6 million in
1985 to more than $30 million in 1994. Business leaders also want to believe
they are having, or at least could have, profound effects on their organizations.
Organizational culture —and whether such culture, seen as a social control sys-
tem, can be managed to enhance organizational performance (O'Reilly, 1989) —
is another topic in which the influence of practical concerns can be seen, as are
the growing number of studies of the effects of pay practices and other human
resource policies on organizational performance (e.g., Huselid, 1995; Pfeffer,
1994; MacDuffie, 1995), studies of organizational change, and research evaluat-
ing organizational interventions designed to enhance performance. Stern and Bar-
ley have argued that being lodged in business schools has "tugged organizational
research toward issues of efficiency and effectiveness and away from larger, sys-
temic issues" (1996, p. 154).
Using organizational culture as an example, Barley, Meyer, and Gash (1988)
demonstrated how to empirically explore the influence of academies on prac-
titioners and vice versa in a study that could serve as a useful model for explora-
tions in other topic areas as well. Barley el al. contrasted two theories of knowl-
edge transmission —"diffusion" and "political":
Whereas diffusion theorists assume that academics frame problems for prac-
titioners, political theorists contend that scholarly endeavors are ultimately de-
fined by the interests of those who dominate society and by whose largess aca-
demics retain the privilege of pursuing research. . . . A more open view of the
relation between academics and practitioners would begin by positing two worlds
that exist as separate but interdependent social systems charactcri/.cd by different
traditions, languages, interests, and norms. Under such conditions, the direction
and degree of influence might vary from issue to issue. (1988, pp. 24-25)
the influence of practitioners necessarily exists for all subjects, the study is instruc-
tive in both its method and its results and is consistent with the claim made here
about the influence of practical concerns on academic discourse, which results,
to some degree, from where the study of organizations is sited.
Abrahamson (1996) has argued for the use of bibliographic indices both over
time and across countries to examine the rise and fall of management fashions —
topics and perspectives that are in or out of favor. For instance, he traced the
rapid rise of interest in quality circles beginning around 1980 and the subsequent
fairly rapid decline in 1984. Abrahamson argued that "fashionable management
techniques must appear both rational (efficient means to important ends) and
progressive (new as well as improved relative to older management techniques)"
(p. 255). lie noted that "intraorganizational contradictions, as well as changes in
their economic and political environments widen certain organizational perfor-
mance gaps creating incipient preferences influencing management fashion de-
mand" (p. 275). The studies of Abrahamson and Barley, Meyer, and Gash (1988)
demonstrate the importance and the feasibility of being more self-reflective as a
field and empirically examining the rise and fall of various ideas and topic areas.
Of course, trends tend to set in motion forces to create their opposite. The
increasing emphasis on productivity, performance, and the interests of organiza-
tions has prompted the development of a critical studies subcliscipline that focuses
on the perspective of those who work in organizations in contrast to those who
own or control them (Stcffy and Grimes, 1986). Thus, for example, Martin
(1992a) wrote about subcultures in organizations and has questioned whether
culture can, in fact, be managed. Other writers have been concerned with the
place of the powerless in organizations arid organization theory, with the explicit
objective of undermining what otherwise would be considered normal or legiti-
mate ways of organizing (e.g., Mumby, 1988). Such writing occasionally focuses
on women and minorities, who often do not fare well in the competition for
salary and positions, particularly compared to their human capital endowments,
and who some argue have been neglected in the writing about organizations (e.g.,
Omi and Winant, 1986; Nkomo, 1992). Still others have argued for a gendered
or gender-based theory of organizations (Mumby and Putnam, 1992; Ferguson,
1984). Much of this literature argues that "theorizing constitutes organizations in
particular ways" (Mumby and Putnam, 1992, p. 465) and, therefore, that science
is an inherently political and subjective activity and should be acknowledged as
such. Although the idea of radical or Marxist organization studies may at first
appear implausible, there are quite important critical studies of social control in
the work place (e.g., Bravcrman, 1974; Edwards, 1979; Burawoy, 1979), of the
neglect of the worker in studies of high-commitment work practices (Graham,
1995), and of the interests and role of social class in organizational and intcrorga-
nizational functioning (e.g., Palmer, 1983).
Yet another phenomenon that may partly derive from where organization
studies has come to be located is the continuing profusion of theories and dimin-
ished consensus on what are the important research questions and directions and
on what arc appropriate research methodologies. Some have bemoaned this ab-
sence of paradigmatic consensus and argued it has created problems for doctoral
Development and Scope of Organization Studies 17
student training (Zammuto and Connolly, 1984), impeded the acquisition of re-
sources and legitimacy (Pfeffcr, 1993), and made it difficult to keep up with the
development of the field (Mone and McKinley, 1993). Webster and Starbuck
argued that "ineffective theories sustain themselves and tend to stabilize a science
in a state of incompetence. . . . Theories about which scientists disagree foster
divergent findings and incomparable studies that claim to be comparable" (1988,
p. 95).
Others have argued that this proliferation of theories and methods is healthy
(Zald, 1993; Van Maanen, 1995b). Burrell and Morgan (1979) identified four
worldviews or paradigms found in organizational analysis: functionalist, interpre-
tive, radical humanist, and radical structuralist. The functionalist paradigm treats
"organization as an aspect of a wider societal system that serves the interests of its
members" (Morgan, 1990, p. 15). The interpretive paradigm challenges the cer-
tainty of the functionalist view, "showing that order in the social world . . . rests
on a precarious, socially-constructed web of symbolic relationships that are con-
tinuously negotiated . . . affirmed or changed" (pp. 18-19). Morgan maintained
that the radical humanist paradigm "is concerned with understanding the way
humans construct a world which they often experience as confining . . . and
with finding ways in which humans can exercise control over their own construc-
tions that allow them to express and develop their nature as human beings" (p.
21). The radical structuralist paradigm, with roots in Marxist theory, emphasizes
the importance of self-generated change and logics of action. Although Morgan
(1983) recognized the inherent incompatibility of these various theoretical para-
digms, he nevertheless argued for the benefits of paradigm diversity:
agendas within fields are played out (e.g., Green and Shapiro, 1994; Parker,
1993). The emphasis in much of this literature is on power and politics. Thus,
for instance, Ilassard noted, "For Kuhn, the everyday reality of science is more
akin to the lifecycle of the political community than to the dictates of formal
logic. . . . Kuhn speaks of discontinuous periods of normative and revolutionary
activity" (1990, p. 220). These writings about the evolution of scientific fields,
always interesting, are particularly relevant for understanding the evolution of
organization studies because research is so often housed in interdisciplinary set-
tings (schools of business, management, or administration) where there are fewer
strong disciplinary anchors to direct activity. For this reason, context probably
matters more, and the future is less easily predicted, than in fields with more
paradigmatic consensus and stronger social norms.
The world that organization theory seeks to analyze and describe has changed in
some important ways. Four of the more significant changes arc (a) the increasing
externali/ation of the employment relation and the development of the "new
employment contract," (b) a change in the size distribution of organizations, with
a comparative growth in the proportion of smaller organizations, (c) the increas-
ing influence of external capital markets on organizational governance and deci-
sion making, and (d) increasing salary inequality within organizations in the
United States, compared both to the past and to other industrialized nations.
These changes, to be documented below, pose some challenges for organization
theory. Most directly, they beg for explanation. If people are working in different
arrangements than in the past, if managerial discretion is now more constrained
by financial markets, if pay is becoming more unequal, and if large organizational
size now seems to be more of a burden and less of a benefit, then organization
theory should have something to say about why these changes have occurred.
Unfortunately, much of organization theory is silent about the structuring of the
employment relation (Pfeffer and Baron, 1988) as well as about the size distribu-
tion of organizations, with population ecology (Mannan and Freeman, 1989) be-
ing a partial exception. Granovetter noted, "One might suppose that the fre-
quently demonstrated importance of size would have stimulated an interest in its
determinants, or at least its marginals; but such discussions are absent from the
sociological literature" (1984, pp. 323-24). Although some work has documented
the increasing influence of the capital markets and the erosion of managerialism
(G. F. Davis and Thompson, 1994), there is, again, little explanation for why
these changes have occurred, particularly in comparison with what has or has not
gone on in other countries. Commenting on the study of wage inequality, Baron
and Pfeffer noted that, although the study of inequality had a long history in
sociology and economics, "remarkably little has been done to bring the firms back
into the study of inequality, particularly in ways that are true to the social and
relational nature of work organizations" (1994, p. 190).
Second, these and other changes in organizational form and functioning
Development and Scope of Organization Studies ig
subcontracting has not: only been used more in traditional ancillary services-
such as cleaning, catering, transport, maintenance and security—. . . but has
also begun to replace the use of conventional employment . . . in a much
broader range of occupational groups such as design engineering, marketing, and
publicity, (p. 849)
Job security is also a thing of the past. Wyatt Company (1993) reported that
72 percent of employers in their 1993 survey had layoffs in the previous three
years, and Louis Harris and Associates (1991) reported that 50 percent of the
surveyed firms had laid off substantial numbers of employees in the previous five
years. The American Management Association reported that in each of the years
from mid-1989 to mid-1994 an average 46.2 percent of the firms reported a work-
force reduction. Moreover, "downsizing tends to be repetitive: on average, two-
thirds of the firms that cut jobs in a given calendar year do so again the following
year" (1994, p. 1). Cappclli (1995), reviewing the evidence on average job tenure,
concluded that average tenure for men had declined. These changes have had
predictable effects on morale and commitment. Using unpublished surveys per-
formed over decades, Cappelli noted that there has been "a sharp decline in
employee attitudes in the 1980s, especially regarding commitment to the em-
ployer" (1995, p. 586). A survey of middle managers (Training, 1992) found that
only 2 percent of middle managers believed that dedication to one's employing
organization was a key to success.
The trend toward externalizing employment has tended to differentially affect
women and minorities. Pfcffer and Baron (1988, p. 270) noted that the growth
rate of employment in business services during the 1974-1984 decade was almost
twice as high for women as for men (136 percent versus 79 percent). Self-
employment in business services tripled over the same period, with women ac-
counting for most of that growth (Howe, 1986). Women are also disproportion-
ately affected by the trend toward part-time employment. Because in the United
States health insurance and pension benefits are tied to employment status, and
because one's likelihood of being covered by union or non-union due process
procedures and the salary one earns is affected by the size of one's employing
organization (Stolzenberg, 1978), the changing conditions of employment have
important social consequences.
The size distribution of organizations has also been shifting, with the trend
being toward a higher proportion of smaller organizations (Granovetter, 1984),
although the size of individual establishments (the units in which work is per-
formed) has been fairly stable since the 1920s. Some of this shift may be ex-
Development and Scope of Organization Studies 21
plained by the changing composition of the economy, in which services now play
a comparatively more important role. In 1982, the proportion of the U.S. work
force employed in services equaled the proportion employed in manufacturing
(Granovetter, 1984, p. 324), the first time this had happened. In general, service
organizations are smaller than manufacturing organizations (p. 325), although the
consolidation in financial services (banks and insurance companies) indicates that
this is not invariably the case. But, even in manufacturing, smaller organizations
are more numerous than in the past. This trend toward smaller size is consistent
with the propensity to contract out tasks and to use more contract labor, reducing
the size of the core organizations. The largest organizations —for example, the
Fortune 500 —have been shedding employees. Christensen reported that the
largest companies, those in the Fortune 500, "reduced their work forces by 2.5
million employees between fiscal years 1983 and 1993" (1995, p. 2), a decrease
of 18 percent. The fact that employment has continued to grow in the economy
as a whole means, by definition, that there has been a shift to smaller organiza-
tions. Castro reported that "the number of people employed full time by Fortune
500 companies has shrunk from 19% of the work force two decades ago to less
than 10% today" (1993, p. 43). Scherer (1980) has reviewed evidence indicating
that economies of scale in production occur at relatively small establishment
sizes, so that the economics of production do not dictate the need for large-size
organizations.
Granovetter's (1984) comparative analysis of size distributions of organiza-
tions in Sweden, the United States, and Japan reported that there were compara-
tively more small organizations in Japan and that Sweden's size distribution of
establishments was comparable to that in the United States. The Japanese manu-
facturing sector has typically operated with an emphasis on building networks of
suppliers and using a great deal of outside contracting, which permitted the core
firms to have fewer employees. To the extent this way of organizing production is
increasingly the case in the U.S. economy, then a trend toward a similar and
smaller size distribution of firms might be expected. Indeed, the emphasis on
specialization, comparative advantage, core competencies, and similar strategies
is consistent with both the changing nature of the employment relationship and
the changing size distribution of organizations that one observes.
The size distribution of organizations is important because size is related to
numerous other organizational characteristics such as having an internal labor
market, formalization, complexity and differentiation in jobs, and having full-time
employees (Kalleberg and Van Buren, 1996, p. 54). Size also affects organiza-
tional rewards, even net of other factors associated with size. Kalleberg and Van
Buren (1996) reported that organizational size was positively related to the receipt
of fringe benefits and to perceived promotion opportunities and was negatively
related to job autonomy. They found that the frequently reported positive relation-
ship between size and earnings was no longer statistically significant when other
mediating factors related to size (such as unionization and the presence of an
internal labor market) were controlled.
A third substantial change in the organizational context has been the growing
power of the external capital markets and the corresponding diminished auton-
22 New Directions for Organization Theory
omy of managers, at least in publicly traded companies. Bcrle and Means (1932)
first noted the issue of the separation of ownership from control and argued that
with widely dispersed share ownership among many comparatively small invest-
ors, managers could operate effectively without oversight. The top executive
largely controlled the selection of the board of directors, which ostensibly was to
oversee this individual (Pfeffer, 1972e; Demb and Neubauer, 1992), and each
individual shareholder was typically too small to make it cost effective to attempt
to intervene in corporate governance. Although there were arguments over the
empirical facts—the extent to which large firms were or were not owner-
controlled (e.g., Zeitlin, 1974)~-the presumption was that at least some substan-
tial fraction of large, publicly traded firms provided their managers with substan-
tial discretion to pursue interests such as growth (Baumol, 1959; Marris, 1967) or
the enjoyment of various perquisites (Williamson, 1964).
In the 1980s, this managerial discretion began to erode substantially. The
change occurred in part because of the development of an active market for
corporate control (Manne, 1965). In this market, assets that arc not being fully
utilized or properly managed by current management are bid away by others who
then manage these assets more efficiently. The market for corporate control has
developed as riskier forms of financing—for example, through junk bonds —have
become more institutionalized and as normative prohibitions on hostile raids or
takeovers have diminished (Hirsch, 1986). Kaplan observed that of the 150 largest
U.S. industrial corporations in 1980, by 1988 some 21.9 percent had been taken
over or merged and another five had gone private in management buyouts, a
situation he correctly characterized as "a very active U.S. takeover market" (1994,
p. 515). Kaplan's study also demonstrated that this change in the context of corpo-
rate control was, to this point, primarily a U.S. phenomenon, with only 2.5 per-
cent of comparably sized Japanese companies being merged during the same
period.
Loss of managerial autonomy results not only from the risk of having the
assets currently controlled by management bid away in a hostile takeover or
tender offer if they are not being effectively employed. It is also the case that
dispersed share ownership itself is a thing of the past, with large pension funds
such as CALPERS (the state of California's pension fund), TIAA-CREF (a pen-
sion fund for college and university faculty), and other large retirement plans as
well as banks and mutual funds now controlling a significant majority of the
shares in many large corporations. These large shareholders have discovered that
simply selling their shares if they are dissatisfied with management is difficult.
Given the large concentration of institutional ownership, market liquidity for
large blocs is possibly limited. There are also the transaction costs (brokerage fees)
associated with securities sales. Consequently, these institutions have become con-
siderably more active in pressing management for desired changes that would
presumably serve shareholder interests. And, particularly for the public pension
funds, run by "outsiders" to the corporate world, there are few normative con-
straints or social ties to reduce the ardor with which such corporate reforms are
pursued. Although such shareholder rebellions are still comparatively rare events,
Development and Scape of Organization Studies 23
they are increasing in frequency and by this very faet have taken on a more taken-
for-granted and institutionalized aspect.
The increased potency of capital market control of the corporation has been
reflected in the changing backgrounds of senior corporate executives —to deal
with financial markets, it is helpful to have a background in finance. Fligstein
(1991, p. 322) argued that analyzing the backgrounds of chief executives helps
one understand both actual power and the perception of functional importance.
Fligstein (1987; 1990) has documented the shift in the backgrounds of chief exec-
utive officers during the twentieth century from generalist entrepreneurs at the
beginning of the period, to those with manufacturing backgrounds, to those with
a sales background. In the current period, most CEOs have a background in
finance.
There have been other important changes in the world of organizations that
also have escaped much analysis in the literature. For instance, there is evidence
that rewards within organizations arc increasingly dispersed, and this dispersion is
much greater in the United States than in other industrialized countries. Abeg-
glen and Stalk (1985, p. 192) showed that the ratio of the salary paid to company
presidents in Japan to the salary paid to newly hired employees has decreased
from a pretax ratio of 110 times in 1927 to 23.6 in 1963, 19 in 1973, and 14.5
by 1980. On an after-tax basis, by 1980 the ratio of president to new employee
compensation in Japan was just 7.5 times. And, this difference in executive com-
pensation cannot be explained by differences in pay levels, for the average level
of pay in Japan in the 1980s was comparable to that in the United States and in
Western Europe. In contrast, Crystal reported that in the United States, "where
[the] typical CEO earned total compensation (excluding perquisites and fringe
benefits) that was around 35 times the pay of an average manufacturing worker
in 1974, a typical CEO today earns pay that is around 120 times that of an
average manufacturing worker and about 150 times that of the average worker in
both manufacturing and service industries" (1991, p. 27). He further noted that
these large gaps were not to be found in Japan, Germany, France, and the United
Kingdom. Even in the United Kingdom, whose business community and econ-
omy is most closely aligned with the United States, the differential between CEO
and average worker pay was under 35 times. Although the topic of wage disper-
sion has occupied a great deal of public attention in the United States, and al-
though it is clear that wage dispersion is produced, at least in part, by forces
operating within organizational labor markets (Baron and Pfcffer, 1994), this sub-
ject has not drawn much theoretical or empirical attention in the organizations
literature.
The changing environment for corporate control, changing si/,e distributions
of firms, changing wage dispersion within organizations, and the profound
changes in the governance of the employment relation have not to this point
played a large role in affecting the research agenda of organization studies. In
some ways, this is because the discipline is frequently almost as unconcerned
about context and history as other related disciplines, such as economics, that
have been deservedly critici/,ed for such theoretical and empirical lacunae. As
24 New Directions for Organization Theory
already noted, organization theory is not only largely housed in U.S. journals, it
is also largely focused on U.S. organizations and ideas. In a comparative study of
the development of organi/ation theory, Guillen argued:
the adoption of models or paradigms of organizational management does not
necessarily follow from their scientific credibility and is not solely determined by
economic and technological factors. . . . Managers in different countries
adopted . . . paradigms in selective ways during the twentieth century, de-
pending on the problems they were facing and such institutional factors as their
mentalities and training, the activities of professional groups, the role of the state,
and the attitude of the workers. (1994, pp. 1-2)
It is also the case that the field has tended to eschew engagement with con-
cerns of public policy. The changing nature of the employment relationship and
its differential effects on women and minorities, the changing structure of wages
in firms, and the increasing influence of shareholder capitalism over what might
be called a stakeholder model all have profound social implications. For the most
part, however, the field has been content to leave public policy debates to econo-
mists and, to a lesser extent, sociologists, even though much of the phenomena
of interest occur in organizations.
This neglect is unfortunate because a comparative— both over time and
across place —research approach has much to recommend it for testing theories
and models more rigorously and helping to discover the extent to which analytical
perspectives arc bound by either period, location, or both. Neglect of policy tends
to keep the field of organization studies on the sidelines in debates about issues
in which it potentially has much to contribute. In thinking about the various
perspectives on organizations to be discussed below, the changing features of orga-
nizations—both those highlighted here and others —should be kept firmly in
mind. One way of evaluating organizational models and theories is to ask to what
extent they provide a productive way of understanding and predicting the evolving
nature of organizations.
2
women or minorities, so that they scored less well on that screening device, then
the test or selection standard had to be validated in order to be employed. Valida-
tion, of course, not only involves determining whether test scores predict subse-
quent performance but can entail selecting more broadly from the applicant pool
(to avoid restriction-of-range problems) and then following those selected to see if
their scores on the test predicted their performance. Moreover, since expectations
for performance arc often self-fulfilling (e.g., Archibald, 1974), a true validation
study must gather test scores when an individual enters the organization and then
keep them secret until they can be compared to performance measures that are
uncontaminatecl by knowledge of the individual's scores. This process is quite
costly and could be undertaken only by comparatively large organizations, and
then only for those jobs that had enough incumbents to make the effort worth-
while (e.g., Howard and Bray, 1988). The practical effect of the Griggs decision
and its vigorous enforcement by the Equal Employment Opportunity Commis-
sion was to make employment testing less important in selection and socially
suspect.
Without reviewing in detail the political and social history of the Great Soci-
ety of President Lyndon Johnson or other social reforms, it seems clear that the
1960s and 1970s were periods in which social interventions to affect the environ-
ment of behavior were seen as important. Early childhood education and devel-
opment was emphasized in the Head Start program, the establishment of Legal
Aid promised to give poorer Americans access to the court system for redress of
problems, and the federal role in education was greatly expanded in an effort to
provide better and fairer educational opportunities in the belief that this would
aid social mobility. By contrast, the current political climate emphasizes individ-
ual choice and responsibility. Subsequent Supreme Court decisions have weak-
ened the requirements for demonstrating the adverse impact of tests or other
screening devices. Legal services for the poor are under attack, and current bud-
geting and legislation seek the devolution of responsibility for education away
from the federal government. This different social context with its focus on the
individual has surely helped shape what organizations scholars study as well as
the perspective they bring to that endeavor.
It is not just the historical context that affects how causes for behavior are
perceived. There is also accumulating evidence that the attribution of causes of
behavior is significantly affected by cultural norms and values. In a set of experi-
ments, Morris and Peng tested the idea that:
still the case that even in the case of the most change, the stability in job satisfac-
tion was statistically significant. They concluded that "satisfaction in 1966 was the
strongest and most significant predictor of 1971 job attitudes. Neither changes in
pay nor changes in job status accounted for nearly as much variance as prior job
attitude" (p. 475). Staw and Ross took this result to indicate that job satisfaction
was a stable individual characteristic, such that some people tended to be happier
than others regardless of their circumstances. Staw and Ross went on to argue
that one possible reason that various organizational interventions (such as job
redesign) were not always as effective as their advocates hoped was that "many
situational interventions may be prone to failure because they must contend with
attitudinal consistency or a tendency for individuals to revert back to their basic-
dispositions" (p. 478).
In a critique of the relative variance claims made by Staw and Ross for dispo-
sitions compared to situations, Johns noted that there was a test-retest correlation
of .84 for both job status and pay (the two situational factors considered in the
study) over the five-year period examined. Consequently, "the situational changes
against which Staw and Ross pitted prior job satisfaction were highly constrained"
(1991, p. 90).
The most obvious problems with this initial study—use of a single-item de-
pendent variable, definition of the personal trait of satisfaction from its stability
over time, which is tautological, and failure to control for a number of alternative
explanations for the results, such as the content of the jobs being performed —
were overcome to varying degrees in subsequent studies. The basic paradigm —
searching for longitudinal correlations in attitudes—remained essentially un-
changed. Thus, for instance, Staw, Bell, and Clausen (1986) used data on person-
ality collected at Berkeley on individuals when they were young to examine job
satisfaction decades later. They found that personality did predict job satisfaction
and that this result held even when the socioeconomic status of the individuals
was statistically controlled. Gerhart (1987) replicated the Staw and Ross study
using a sample of younger workers (who would presumably be more likely to
experience more change in job conditions). He attempted to control for differ-
ences in job characteristics by assigning job characteristic scores to occupations
on the basis of their descriptions in the Dictionary of Occupational Titles and by
using those items from the Job Characteristics Inventory (Sims, Szilagyi, and Kel-
ler, 1976) that measured job complexity. He found that although there were im-
portant effects of job characteristics on job attitudes, the statistical significance of
the cross-time stability in job satisfaction remained.
One issue in all of these studies is what stability in job satisfaction scores
actually says about the existence of dispositions. Gutek and Winter (1992) argued
that this apparent stability was not necessarily an indicator of some stable underly-
ing trait but rather possibly an artifact of response-shift bias (Howard and Dailey,
1979). Response-shift bias raises the possibility that there is a change in the frame
of reference used by the respondent in answering questions about job satisfaction
so that the stability in job attitudes is illusory. Gutek and Winter wrote:
overall level differences that reflect that some work environments are generally
more satisfying than others" (Watson and Slack, 1993, p. 182).
Any theory emphasizing the importance of dispositions must necessarily an-
swer the question of where such dispositions originate. One possible answer is
genetics, and a series of studies of twins has sought to demonstrate the heritability
of both job attitudes and vocational interests. These studies have capitalized on
the availability of data gathered on twins, some of whom were separated at birth
and raised apart—something that is important in order to be able to empirically
distinguish between genetic heredity and environment. Arvey, Bouchard, Segal,
and Abraham (1989) estimated that about 30 percent of the observed variation in
general job satisfaction could be attributed to inherited or genetic factors. Lykken,
Bouchard, McGue, and Tellegen (1993), studying vocational and recreational
interests in pairs of twins, found that about 50 percent of the variation in interests
was associated with genetic variation. They also found a higher degree of correla-
tion in the monozygotic twins than for the dyzygotic twins, the latter having less
genetic material in common. Note that the amount of heritability implied by
these studies is substantial —- in some cases exceeding 50 percent.
There have been a number of methodological critiques of the twin studies.
Cropanzano and James (1990) noted that if the twins were raised in similar envi-
ronments—a plausible assumption —comparing environmental with inherited in-
fluences would not be valid. But perhaps the biggest issue is that, "absent infor-
mation regarding how the process works, one wonders what is inherited" (Judge,
1992, p. 44, emphasis in original). There have been numerous studies demonstra-
ting the likelihood that intelligence is inherited, at least in part, and has a genetic
component (Bouchard and McGue, 1981; Seligman, 1992; Hernstein and Mur-
ray, 1994). Intelligence, moreover, has been shown to affect numerous life out-
comes such as career success (O'Reilly and Chatman, 1994). The question posed
by the twin studies is whether the observed relationship between job satisfaction
and vocational interest scores across sets of twins is the consequence of the herita-
bility of job affect and career interests or of the heritability of intelligence and
the effect of intelligence on career interests and job satisfaction (see also Judge,
1992). This distinction, however, is not necessarily critical to those who argue
that the heritability of traits and their stability and existence are important, not
the particular mechanisms or even the particular traits that are inherited. Thus,
for instance, Lykken et al. noted, "the experiences people seek, and the effect of
those experiences on their developing interests, are influenced by traits of phy-
sique, aptitude, and temperament . . . that are themselves substantially geneti-
cally influenced" (1993, p. 658).
One critical problem faced by research on the dispositional bases of organiza-
tional behavior has been to identify not only the source of dispositions but also a
consistent, replicable set of meaningful dispositions. As noted previously, the work
on the stability of satisfaction over time led naturally to an interest in the possibil-
ity that one important individual trait was a propensity to be happy or unhappy.
This corresponded nicely with work in social psychology on positive and negative
affect and has resulted in a number of empirical studies. Watson, Clark, and
Tellegen (1988) reported on the development of scales for positive and negative
32 New Directions for Organization Theory'
affect and evidence that these are, in fact, two orthogonal dimensions. Costa and
McCrac (1988) reported six-year test-retest correlations of 0.82 for positive affect
and 0.83 for negative affect, providing evidence of the stability of this disposition
over time.
Levin and Stokes (1989) argued and provided evidence that negative affcctiv-
ity was related to measures of job satisfaction. Their experimental study manipu-
lated task characteristics by giving some subjects the job of rating graduate stu-
dent applications and others the task of copying transcripts. They selected subjects
that were cither in the highest or lowest quartile of the scale of negative affcctivity.
Their empirical results indicated that although task condition explained some 53
percent of the variation in job satisfaction while negative affectivity accounted for
only 4 percent, even in the two very different task conditions the effect of negative
affectivity was statistically significant. Watson and Slack (.1993) studied eighty-two
employees at Southern Methodist University, measuring them initially on scales
assessing positive affect and negative affect. Subsequently, subjects were retcsted
on these personality measures and also completed job satisfaction and job change
measures. Although there was significant attrition in the group, through both
voluntary turnover and firing, the initial measures of affect were unrelated to
whether or not the subjects remained employed at SMU. Measures of tempera-
ment showed good test-retest consistency, with correlations ranging from 0.63 to
0.74. Also, initial measures of positive and negative affect were correlated with
job satisfaction almost two years later, and, in fact, the correlations between affect
at the time of the initial measurement and subsequent satisfaction were of about
the same magnitude as contemporaneous correlations between the dispositions
and job satisfaction when measured at time 2. Moreover, even when (perceptual)
measures of job characteristics were partialed out in the analysis, the effect of
positive and negative affect remained.
Of course, positive and negative affect are not the only critical dimensions of
personality. Another advance made in the study of dispositions has been conver-
gence on the importance of five core dimensions of personality—the so-called
Big Five (e.g., Barrick and Mount, 1991; Digman, 1990; L. R. Goldberg, 1990).
The Big Five factors are: cxtraversion, agrccablcncss, conscientiousness, emo-
tional stability, and openness to experience (Barrick and Mount, 1993, p. 111).
This is an important advance because one of the criticisms of dispositional re-
search has been that it was atheorctical and tautological. For instance, Judge
and Ilulin commented on the proliferation of "empirical studies linking traits
to outcomes, with often only tangential relations to a well-developed theoretical
framework" (1993, p. 389). The search for personality dimensions has at times
involved finding some regularity in behavior and then naming that regularity a
personality dimension; at other times it has involved factor analyses of question-
naire responses, which data has been used to empirically uncover dispositions.
These factor-analytic solutions were not always stable across subject populations,
and the search for dispositions was criticized for attempting to capitalize on em-
pirical regularities.
O'Reilly and Chatman (1994) have noted that although there has been a
renewed interest in the effects of individual characteristics on organizational be-
Understanding the Causes of Behavior 33
Although there has been a plethora of articles and substantially increased interest
in exploring the effect of individual traits on behavior in organizations, this litera-
ture is not without its problems. First, there remains ambiguity about the defini-
tion of some of the traits, including even the Big Five. Barrick and Mount ac-
knowledged that "some researchers have reservations about the five-factor model,
particularly the imprecise specification of these dimensions" (1991, p. 3). These
reservations flow logically from the fact that "while there is general agreement
among researchers concerning the number of factors, there is some disagreement
about their precise meaning" (p. 3). For instance, the first trait is extraversion/
introversion, associated with the characteristics of being sociable, talkative, and
assertive (p. 111). However, liogan (1986) has interpreted this dimension as con-
sisting of two components, ambition and sociability. Extraversion, ambition, socia-
bility, and talkativeness are not necessarily conceptually identical. Similar impre-
cision exists for virtually all of the traits. For instance, conscientiousness has also
been called conformity or dependability and has been referred to as the will to
achieve because of its association with various educational achievement measures:
"As the disparity in labels suggests, there is some disagreement regarding the es-
sence of this dimension" (Barrick and Mount, 1991, p. 4).
Similar problems affect the interpretation of positive and negative affect,
which have been most often measured and theorized as if they were statistically
independent dimensions. But the independence between the constructs seems to
depend on measurement and methodological issues (Chamberlain, 1988; Diener,
1990). Diener concluded that "there is not replieable evidence across samples
and methods that positive and negative affect are . . . unrelated" (1990, p. 14).
Second, although some studies have measured the stability of dispositions
across time, those dealing with positive and negative affect, for example, such
concern about ensuring that stable traits are identified and validated is not neces-
sarily the norm. For example, several authors have employed the inaugural ad-
dresses of U.S. presidents to measure presidential personality (e.g., House, Span-
gler, and Wyocke, 1991; Spangler and House, 1991; Winter, 1987). Winter coded
inaugural addresses for achievement, affiliation, and power-motive imagery. How-
ever, he chose to code a single speech-—the first inaugural address. Although a
number of the presidents were rcclccted and therefore gave second inaugural
speeches, and since many had been in public life for years and so had given
speeches that could also have been analyzed, there was no attempt in any of
the studies of the effects of presidential personality to establish that presidential
personality, as a reasonably stable underlying construct, had been measured by
using the same method to assess consistency over time.
Indeed, Winter's own data permit one to explore an alternative interpretation
for his results —namely, that the imagery in the speeches reflects not the personal-
ity of the president but rather the political ideology of the president's party. Win-
ter (1987) reported the scores for the U.S. presidents he studied. Considering only
those from the twentieth century, Democratic presidents scored, on average, 62.8
on the standardized power-imagery measure while Republican presidents scored
Understanding the Causes of Behavior 35
49.5. Democrats scored 60.1 on affiliation while Republicans scored 51. Thus,
across political parties there are differences that provide a plausible explanation
for the differences in imagery observed, without relying on the construct of per-
sonality.
Nor is there increased attention to measuring the stability of dispositions in
more recent research, conducted long after this problem was discussed (Davis-
Blake and Pfeffer, 1989). For example, in a study of the effect of the individual
characteristic of cooperativeness and of being in a more cooperative culture on
cooperative behavior, Chatman and Barsade (1995) measured individual coopera-
tiveness at a single point in time, using a scale that had not been shown to be
stable over time.
Third, there is evidence that dispositions or elements of personality appear to
change in response to learning, experience, and exposure to different situations.
House, Howard, and Walker (1991) found that the level of managerial responsi-
bility AT&T employees had attained at one point in time had a statistically sig-
nificant effect on the will to manage at a later time. Kohn and Schooler (1978;
1982) reported that individual traits such as self-direction and tolerance for uncer-
tainty changed depending on the particular jobs that people held over time.
Fourth, the definition of dispositions is frequently imprecise with respect to
the stability implied in the definition or to the distinction between dispositions
and other psychological constructs. Thus, House, Shane, and Herold wrote, "Dis-
positions generally are viewed as tendencies to respond to situations . . . in a
particular, predetermined manner. Furthermore such tendencies may vary in
their temporal stability, their activation state, and their usefulness in explaining
behavior under different conditions" (1996, p. 205). That article also noted that
while personality characteristics were the most stable dispositions, "needs, moti-
vates, preferences, and attitudes . . . need not be viewed as stable" (p. 205). This
muddies the distinction between individual dispositions and other psychological
states, including mood, which are readily induced (e.g., Isen and Levin, 1972;
Brief, Butcher, and Roberson, 1995).
Fifth, many of the effects of dispositions, particularly on outcomes of organi-
zational interest, are incredibly weak if not nonexistent. O'Reilly and Roberts,
using a sample of 578 officers and enlisted men in a high-technology navy unit,
found "no significant relationship between intrinsic traits and job satisfaction"
(1975, p. 148), once factors such as rank and tenure with the navy were con-
trolled. Barrick and Mount (1991) performed a meta-analysis of the effect of the
Big Five personality traits on job performance, reviewing 117 studies conducted
between 1952 and 1988 that covered some 24,000 individuals. They reported
that the magnitudes of the estimated true correlations between these personality
dimensions and job proficiency, training proficiency, and various career outcome
measures such as salary, turnover, and tenure were less than 0.10. Tell, Jackson,
and Rothstein's (1991) meta-analysis of the personality-job performance relation-
ship in ninety-seven field studies found that the average corrected correlation
between personality and job performance was 0.19. Hough et al. (1990), examin-
ing the validity of six personality traits in a sample of over 9,000 individuals in
the U.S. military, found predictive validities for a variety of job-related criteria in
36 New Directions for Organization Theory
the 0.20s. The average correlation between the traits and both technical profi-
ciency and general soldiering proficiency was 0.04. The outcomes best predicted
by the traits were physical fitness and military bearing. These reviews indicate
that dispositions invariably account for 4 percent or less of the variation in out-
comes such as job performance.
Examination of some specific studies illustrates the point that even when
statistically significant correlations with personality dimensions are found, the ef-
fect sizes are often minute. Bluen, Barling, and Burns's (1990) study of the rela-
tionship between personality and the number of policies sold by life insurance
brokers found a correlation 0.09 with impatience-irritability and 0.18 with
achievement strivings. The same study showed that experience (organizational
tenure) correlated 0.39 with the number of policies sold, demonstrating that work
experience was a better predictor of current sales than personality. George's (1989;
1991) studies of the effects of positive and negative affectivity found that absentee-
ism from work correlated —0.10 with positive affectivity and 0.08 with negative
affectivity. She also reported that positive affectivity was correlated 0.10 with a
measure of altruism, 0.02 with indicators of customer service, and 0.00 with sales
performance. Her studies found that state or mood had a stronger effect than
dispositions on the outcomes assessed, but she reported that the strongest effects
on absence were job tenure, age, being paid on commission, and full- or part-
time employment status, all situational factors.
Sixth, many of the studies of dispositions make virtually no effort to control
for plausible situational factors that might also explain the results. An important
consequence of failure to control for situational causes is that the unique effects
of situations and dispositions become impossible to untangle, and there is the
potential for misattributing situational effects to dispositional causes.
Both self-selection and organizational selection result in the nonrandom as-
signment of individuals to organizational conditions (Schneider, 1987). In this
sense, much research on individual attitudes and behaviors in organizational set-
tings consists of quasi-experiments. In these quasi-experiments, modeling only dis-
positional determinants of behavior can lead to biased estimates of effects. Achen's
(1986) book on the problems involved in making causal inferences from data
generated by quasi-experiments discussed this point at length. He demonstrated
that "an accurate assessment of a quasi-experiment depends on explicit modeling
of both the behavioral outcome of the experiment and the assignment to treat-
ment groups" (p. 37). Achen showed that, unless the assignment process is mod-
eled, the results of quasi-experiments are frequently worthless. In organizations
research, modeling the assignment process requires estimating the effects of orga-
nizational characteristics (such as the selection and reward system) that both
cause individuals to be in particular organizations and that also cause many of
the dependent variables being studied. Cropanzano and James (1990) have made
a similar point in their critique of the study of twins.
It should be noted that not every study of individual dispositions suffers from
all of these problems. For instance, Judge and Hulin made a carefully drawn
distinction between "affective disposition, defined as the tendency to respond gen-
erally to the environment in an affect-based manner, and subjective well-being,
Understanding the Causes of Behavior 37
the level of overall happiness and satisfaction an individual has with his or her
life" (1993, p. 388) and were quite precise in their measurement models. O'Reilly
and Chatman (1994) also have taken care to measure the traits carefully, ensure
their validity, and control for at least some situational factors. But the problems
are pervasive enough in the literature so that readers should keep them in mind
in evaluating this research.
There are a number of authors (e.g., George, 1992; House et al., 1996) who argue
that organization studies should accept the interactionist perspective—situations,
individual traits, and their interactions are all important—and stop debating the
point. George, for instance, argued that "rather than mimicking the debate that
raged in psychology over these same issues, it will probably be much more fruitful
to try to understand the effects of person factors, situational factors, their various
types of interactions, and the processes governing these interactions on organiza-
tional phenomena" (1992, p. 192). House ct al. referred to the person-situation
controversy as a "pseudo issue" and suggested that "it is more meaningful to ask
how dispositional variables and organizational variables interact in evoking behav-
ior" (1996, p. 220). Pervin (1985) and Carson (1989) also expressed exasperation
with the continuing disputes about whether individual traits mattered or which
were more important in general. The argument seems to be that the debate di-
verts time and attention away from advancing knowledge about organizational
behavior.
The continuing controversy does, however, potentially have some positive
benefit. First, the very fact of controversy, debate, and the passions aroused
thereby provide attention, energy, and emphasis on this particular subset of the
field. Thus, somewhat ironically, critiques of the dispositionalist perspective in
organization studies (e.g., Davis-Blake and Pfeffer, 1989), and the responses they
generated (e.g., House et al., 1996), have probably stimulated more and better
work on this topic than might otherwise have occurred (see also Kenrick and
Funder, 1988). As Mone and McKinley (1993) have argued, organization studies,
in part because of its pre-paradigmatie nature, values uniqueness and things that
are interesting possibly more than many scientific fields do. lb the extent this is
the case, the very controversy of the debate provides a measure of how interesting
and novel the subject is and thus encourages more research activity.
Second, there is little doubt that critiques of the dispositionalist view have
stimulated the substantive development of the study of individual traits. Without
the critiques concerning the conceptual ambiguity of traits, methodological issues
including failure to control for relevant situational influences and to use appro-
priate methods, and the very meaning of dispositions, one might question whether
the field would have seen the advances in method and conceptual development
witnessed particularly over the past several years.
The point about the relevance and usefulness of controversy is a general one
for the field of organization studies, for this particular controversy is not the only
one that has engaged the field. The fundamental issue is whether knowledge will
be better developed by accommodating a number of points of view or whether
the very absence of accommodation, particularly initially, stimulates deeper and
Understanding the Causes of Behavior 41
more insightful thinking and analysis, to the ultimate benefit of the development
of both theory and data.
Zajonc (1980, pp. 202—203) made much the same point in discussing research
on social perception and social cognition: "Social perception and social cognition
create a social reality by affecting both the perceiver and the objects perceived.
. . . Thus, I have argued, social perception requires equal attention on both sides:
on the side of the observer or pcrccivcr, and on the side of the objects perceived."
Although many models of behavior are possible, here I identify and describe
five of the most important models of action and choice that are prominent in
organization studies. In each instance, we want to examine the underlying as-
42
Five Models of Behavior 43
sumptions, the research directions these models have produced, and their impli-
cations for understanding organizations. The five models are (1) the economic
model; (2) the social model; (3) the retrospectively rational model; (4) the moral
model; and (5) a cognitive, interpretive model.
There is, of course, a first question as to what extent these models should be
evaluated by the veracity of their assumptions or by the empirical confirmation of
their predictions. Friedman (1953) articulated a position that has often been re-
peated: models (and he was referring to economic models) should not be evalu-
ated on the validity of their assumptions but rather on the validity of the predic-
tions generated by such models. A variant of this position is the natural-selection
or competition argument. This argument maintains that even though, in a par-
ticular instance, a firm or individual may not engage in rational or profit-
maximizing behavior, over time, competitive selection pressures will ensure that
these decisions, individuals, or firms diminish in relative frequency (Becker,
1962). In this sense, there is no need to show how various adaptive, efficient
practices arise, because "it is implicitly assumed that inefficient ones will have
failed the test of the marketplace" (Granovetter, 1985, p. 30).
Although this epistemological position is still generally the rule in economic
analysis, it has been subjected to numerous criticisms for the logical fallacies it
contains (Samuelson, 1963; Simon, 1963). Blaug (1980) has offered a critique of
natural selection, adaptive reasoning. Boland (1979) has noted that the basic pur-
pose of a theory is to explain, and to rely solely on whether or not the predictions
are valid is foolish. "When predictions prove to be valid, we do not know why,
and hence are unable to foretell under what conditions they will continue to hold
or fail, or may need to be adapted" (Etzioni, 1988, p. 17).
One argument is that unrealistic models or theories may offer an advantage
in simplicity or parsimony over their more realistic but complex competitors. Si-
mon discussed the trade-off between theoretical simplicity and the assumptions
that were required by the theory:
Occam's Razor has a double edge. . . . Occam understood his rule as recom-
mending theories that make no more assumptions than necessary to account for
the phenomena. . . . A theory of profit or utility maximization can be stated
more briefly than a satisficing theory. . . . But the former makes much stronger
assumptions than the latter about the human cognitive system. . . . The two
edges of the razor cut in opposite directions. (1979, p. 495)
The arguments over the bases for evaluating models of behavior help to dis-
tinguish between the economic model and the others reviewed below, most of
which have been created to more accurately and completely capture empirical
reality, even at some cost in simplicity or elegance.
Resolving this issue about realism versus parsimony involves trade-offs, and
different fields and different theorists make those trade-offs differently. A parsimo-
nious and, in economics, mathematically elegant theory simplifies or neglects
institutional details that vary by time, place, and circumstance, but it may do so
at the expense of a more complete model that can capture all of the institutional
complexity. Ronald Coase, writing about the old institutional economics, made
44 New Directions for Organization Theory
his choice on the side of theoretical elegance: "Without a theory they had noth-
ing to pass on except a mass of descriptive material waiting for a theory, or a fire"
(1983, p. 230). As Scott has noted, "the battle between the particular and the
general, between the temporal and the timeless, is one that . . . theorists con-
tinue to confront" (1995, p. 5).
has contrasted the idea of efficiency with that of power and has vigorously main-
tained that power cannot account for organizational outcomes. Organizational
design, practices governing the employment relation, and patterns of interorgani-
zational relations that survive over time are all efficient, according to Williamson,
and do not reflect the operation of political interests and relative potency of those
interests. This assertion means that literature assessing the effects of power and
interests on these organizational arrangements provides a partial test of the eco-
nomic model of organizations.
Second, virtually all economic models proceed from a theoretical apparatus
invoking methodological individualism. That means simply that institutions or
organizations are seen as aggregations of individual preferences and actions or
at times as a nexus of contracts and agreements (Jensen and Meckling, 1976).
Organizations are viewed as having no substantive reality apart from this function
of aggregating preferences or agreements. Williamson argued that organizations
are simply another type of "contractual instrument, a continuation of market rela-
tions, by other means" (1991, p. 162; emphasis in original).
Third, economic models emphasize comprehensiveness and often proceed
from an assumption of equilibrium. Coase has argued that one reason economic
analysis has been successful in moving into related social sciences is that the
analysis of systems equilibria focuses on systemic interdepenclencies and is a mode
of analysis "more likely to uncover the basic interrelationships within a social
system" (1978, p. 209). The focus on equilibrium conditions and the assumptions
of market competition tend to assure a more deterministic outcome from the
analysis. In emphasizing equilibrium conditions, modern economics demonstrates
less concern with "path-dependence," or the idea that what has happened in the
past is important for understanding the present and predicting the future. Both
history and process are less important in this form of theorizing, which Jacoby
has called "timeless and placeless" (1990, p. 32), with no emphasis on how the
economy acquired its features and how these vary.
Although many social science theories share an assumption that individuals
pursue self-interest, a number of the economic theories that have penetrated orga-
nization studies, such as transaction cost economics and agency theory, take the
idea of self-interest one step farther and emphasize opportunism, or self-interest
pursued with guile and deceit (Ghoshal and Moran, 1996). As Ghoshal and
Moran noted, there is an important distinction between opportunism and self-
interested behavior:
their personal use" and that "cooperative effort is plagued by opportunities for
malfeasant behaviour" (Nilakant and Rao, 1994, p. 650). In this sense, many
economic theories differ from other social sciences in their assumptions about the
extent to which human behavior is strictly instrumental and unconstrained by
morals or values, except when behavior may be discovered and have consequent
adverse effects on reputational resources.
The presumption of opportunism, an "extreme caricature" (Milgrom and
Roberts, 1992, p. 42) of human behavior, is critical to at least some variants of
the economic model. For instance, Williamson noted that "but for opportunism,
most forms of complex contracting and hierarchy vanish" (1993, p. 97). Ghoshal
and Moran, summarizing transaction cost theory, noted that "opportunism — the
seeking of self-interest with guile —is the ultimate cause for the failure of markets
and for the existence of organizations" (1996, p. 17).
Because people are presumed to act in a self-interested fashion to maximize
their utilities, and because individuals are presumed to have a disutility for ex-
pending effort, economic models invariably view employees as effort-averse, un-
likely to do what the organization wants and needs without some form of incen-
tive, sanction, or combination of both. Baron has noted:
The image of the worker in these models is somewhat akin to Newton's first law
of motion: employees remain in a state of rest unless compelled to change that
state by a stronger force impressed upon them— namely, an optimal labor con-
tract. Various incentive features . . . are claimed to provide forms of insurance
that overcome workers' reluctance to work. (1988, p. 494)
rule (the real cost of engaging in some action is the loss of benefits available from
pursuing the next best course of action) (Larrick, Nisbett, and Morgan, 1993).
Larrick et al. reported that "people who use these rules are more likely to have
more successful life outcomes . . . college seniors . . . had higher grade point
averages . . . and faculty . . . had higher salaries" (1993, p. 345). Frank, Gilo-
vich, and Regan (1993) summarized studies indicating that individuals with eco-
nomics training tended to defect more frequently in prisoner's dilemma experi-
ments and tended to give less to charity. However, Arkes and Blumer (1985)
reported no effect of previous courses in economics on not considering sunk costs
in decision making, and Fischhoff (1982) has also concluded that many attempts
to train people to avoid decision biases were unsuccessful. The two sets of results
together suggest that behaving on the basis of classical economic assumptions
may provide individual benefit, although whether or not it provides general social
or even organizational benefit is open to question and depends on the amount of
voluntary cooperation required for system effectiveness.
Because of the emphasis on methodological individualism and the efficiency
of market-mediated transactions, the question naturally occurs as to how and why
institutions, such as organizations, arise at all. One approach to answering this
question is provided by transaction cost economics (Williamson, 1975). As re-
viewed by Williamson, the basic elements of the model are (1) the transaction is
the basic unit of analysis; (2) transactions vary with respect to their frequency,
uncertainty, and asset specificity, which affects how easy it is to redeploy the assets
to other uses; (3) there are a set of generic modes of governance, including mar-
kets, hierarchies, and intermediate forms; and (4) the basic prediction is that
"transactions, which differ in their attributes, are aligned with governance struc-
tures, which differ in their costs and competence" (1995, p. 27) in a way so as to
minimize transaction costs.
These governance problems arise because of bounded rationality and uncer-
tainty, which make the future impossible to foresee perfectly and make con-
tracting for all possible future states of the world unfeasible. One might wonder
why the parties involved don't simply handle contingencies as they arise. The
answer, of course, is that each individual is presumed to be a utility-maximizing
person interested in attaining his or her own interests and so cannot be trusted.
This particular economic model of behavior has within it assumptions of the
same sort of untrustworthy behavior typical of the genre —in this case, opportun-
ism. "By opportunism, I mean self-interest seeking with guile. This includes, but
is scarcely limited to, more blatant forms such as lying, stealing, and cheating.
Opportunism more often involves subtle forms of deceit" (Williamson, 1985,
p. 47).
Transaction cost models also provide an account for the presence of hierar-
chy in organizations, where the degree of hierarchy is defined as the extent of
centralization of power and control (Williamson, 1985). Under the usual adaptive
logic, hierarchy exists because of its efficiency: "the least hierarchical modes, in
both contracting and decision-making respects . . . have the worst efficiency
properties" (Williamson, 1985, p. 231).
If the reader is having trouble accepting the behavioral premises or assump-
48 New Directions for Organization Theory
tions of these models, it is not surprising. Not only are these models at variance
with much of social psychological theory—for instance, theories emphasizing in-
trinsic motivation (Lcpper and Greene, 1975; Deci, 1975) —they also contradict
much current management practice that decrnphasizes hierarchy, individual
monetary incentives, and control. Furthermore, the economic model of behavior
presents an extremely unflattering portrait of human behavior. "Typically, such
organizational economic actors seek personal wealth, status, leisure, or the like.
. . . Accordingly, their behavior as potentially shirking or opportunistic agents can
be curbed by vigilant monitoring, together with incentive schemes based around
money, promotion, negative sanctions, and the like" (Donaldson, 1990, pp. 371-
372). To the extent such theories of behavior, when implemented in actual orga-
nizational policies and practices, produce the very behavior expected (i.e., be-
come self-fulfilling prophecies), to the extent such theories of behavior are grow-
ing in importance in the social sciences, and to the extent such theories are at
variance with basic ideas in sociology, social psychology, and organization theory,
it seems clear that this is a substantial and important arena in which these social
sciences need to challenge these models of behavior.
pirical papers to nonempirical papers is 1:14 (Barney and Ouchi, 1986)" (1995,
p. 1226).
When data are employed, because ot the belief in the theory, almost any
massaging of the data to get it to conform to the theory is acceptable. The data
used are frequently previously collected for other purposes, and assumptions and
variables are added on an ad hoe basis (Etzioni, 1988, p. 18) to improve the fit
between theory and data. This makes it hazardous to evaluate the validity of the
models' predictions: "When we defend the maximization paradigm by pointing to
the similarities between its predictions and observed behavior, we often overlook
the fact that the empirical 'success' of many economic models is principally de-
rived from accommodating adjustments in complementary hypotheses" (Cross,
1983, p. 3).
The assumptions of adaptive efficiency characteristic of these models almost
invariably lead to tautological reasoning. Practices are presumed to be efficient
because of their very existence —if they were inefficient they would disappear—
and thus the logic of economic science as it is practiced is, given a particular
empirical observation, to derive a proof that demonstrates the efficiency properties
of what has been observed.
As one example, consider the analysis of seniority-based wage systems in orga-
nizations (Lazear, 1979; 1981). In most organizations and for many jobs, there is
a positive relationship between experience and earnings. But, it is not likely in
all organizations and jobs for which this relationship holds that an individual's
productivity goes up linearally with experience. So, the question is, Why not pay
workers each period on the basis of their productivity for that period, thereby
inducing effort to attain those contingent rewards? What Lazear demonstrates is
that a payment scheme that rewards employees less than their marginal product
when they are young and more when they are old can motivate effort because
there is now an incentive for the more senior workers not to shirk—if they are
discovered and fired, they lose the deferred compensation inherent in being paid
more than they are producing. And, younger workers are motivated to work hard
and stay with the firm so they can enjoy the benefits of higher earnings when
they have more experience. Since this story begins with an empirical fact and
then makes economic sense of that fact, it does not do a very satisfactory job of
accounting for the recent trends away from this sort of payment system—although
1 have no doubt that such a change could be readily accounted for by an
efficiency-based account if economic theorists were called to do so. This tendency
toward tautology has been noted by Sen (1977), Arrow (1982), and Latis (1976),
among others.
Another problem is that the methodological individualism characteristic of
these models either ignores organizations and institutions almost completely or
treats them as a residual category required by some form of market failure or
contracting problem. Clegg noted that "economic approaches to organizations
. . . operate with a particular set of a priori values: these are that organizations
are an aberration from a more natural form of economic activity. This more natu-
ral form is that of exchange on the market" (1990, p. 61). Simon, who also won
50 New Directions for Organization Theory
the Nobel prize in economics, has critiqued even the new institutional economics
for its relative neglect of organizations:
A fundamental feature of the new institutional economics is that it retains the
centrality of markets and exchanges. All phenomena are to he explained by
translating them into (and deriving them from) market transactions based upon
negotiated contracts, for example in which employers become "principals" and
employees become "agents." . . . The new institutional economics is wholly
compatible with and conservative of neoclassical theory. . . . In general, the
new institutional economics has not drawn heavily from the empirical work in
organizations and decision-making for its auxiliary assumptions. . . . Until . . .
the existing literature on organizations and decision making [is] taken into ac-
count, the new institutional economics and related approaches are acts of faith,
or perhaps of piety. (1991, pp. 26-27; emphasis added)
rational for its strongest members. Indeed, why the best-performing corporations
. . . maintain their membership is, at least to economists viewing firms as ratio-
nal utility maximizers, perhaps the greatest puzzle posed by the keiretsu phenom-
enon. . . . There is a collective logic to the keiretsu phenomenon that, in our
view, is not reducible to rational optimizing on the part of individual firms.
(1996, p. 86)
But perhaps the most compelling critique of the economic model, at least as
it has been applied to understanding organizations, is its difficult)' in making
unique predictions about empirical results. Consider first an issue that is well
within the scope of the new institutional economics of Williamson and others and
is a prominent feature of the new organizational landscape —should an economic
exchange be incorporated inside the organization's boundaries or left to market-
mediated transactions? Given the significant change over time in the extent to
which the employment relation has been internalized and the change in the size
distribution of organizations that has occurred, in part, because of the more exten-
sive use of contracting arrangements rather than vertical integration, one might
think that economic analysis would provide a compelling answer as to why labor
was first internalized in the employment relation as market-mediated transactions
were supplanted and now is being increasingly externalized once more.
52 New Directions for Organization Theory
But it doesn't. Davis-Blake and Uzzi, studying the use of temporary help
and independent contractors, concluded that "the use of externalized workers
is determined by multiple factors and is affected by the costs and feasibility of
externalization, by the structure of the organization, and by the interests of exter-
nal groups" (1993, pp. 216-217). After reviewing the arguments of Williamson
(1975) and others for the development of internal labor market arrangements that
substituted for spot labor markets, Pfeffer and Baron (1988, p. 263) concluded:
One problem is that one of the crucial variables the theory employs to predict
whether transactions are better left to the market or incorporated inside the orga-
nization, asset specificity, is arguably as much a consequence of the decisions
about how to organize as it is their cause. A firm that has decided to internalize
its employment will be much more likely to engage in training and to organize
jobs on the presumption of worker continuity; a firm that has decided to rely on
temporary help and contract labor will invest less in training and will find ways
to organize production so that long-term attachments by workers are unnecessary.
A quick consideration of the various ways in which work is organized in the
automobile industry provides an example of this. Moreover, the theory docs not
anticipate the empirical reality of many firms that organize transactions both
through markets and hierarchies. For instance, most food franchisers have both
company-owned stores (hierarchy) and franchised operations (market-mediated re-
lations). Except as a transitional state (which is not the case), the economic mod-
els of organizations would predict either one or the other arrangement (but not
both) in a particular case depending on asset specificity and other contracting
issues.
Moreover, there is evidence that both institutionalization and power theories
of organizations occasionally are more successful than some presumption of ratio-
nality or efficiency in explaining both organizational structure and employment
practices. Pfeffer and Cohen (1984) found that, controlling for other factors in-
cluding size and technology, the use of internal labor market practices was posi-
tively related to the presence of a personnel department and negatively related to
the extent of unionization. Unions and personnel professionals are both potent
political groups in organizations. Baron, Dobbin, and Jennings (1986) argued that
modern human resource management practices, implemented under government
pressure during World War II, were maintained after the war in part through the
activities of personnel professionals and with the support of personnel associa-
tions. Internal labor market arrangements and modern personnel practices obvi-
ously provide a more important and central role for the personnel function in
organizations, and it is therefore not surprising that these innovations in personnel
Five Models of Behavior 53
of factors other than the validity of its assumptions or its empirical support. Its
mathematical elegance provides status and makes it less accessible to the un-
trained, which lends it prestige. How can something carry prestige if anyone can
do it? Conversely, some technique or line of reasoning accessible and understood
only by the few carries the value imparted by scarcity (Cialdini, 1988). The para-
digmatic consensus born out of a way of thinking about problems provides the
field with greater potential for unity of thought and action, always a major advan-
tage in political contests (Pfeffer, 1992).
Furthermore, economists have taken a number of strategic behaviors to en-
force and maintain theoretieal hegemony (R. Parker, 1993). S. Cole has argued
that "one of the primary mechanisms through which consensus is maintained is
the practice of vesting authority in elites" (1983, p. 137). R. Parker (1993) has
documented the dominance by graduates of elite schools in the discipline of
economics. And those who do not conform to the orthodoxy of the discipline are
excluded from its most prestigious positions, regardless of the importance of their
ideas or their general scholarly reputation. The case of Robert Reich, secretary of
labor in the Clinton administration, offers a good case in point:
Although the general approach of Mr. Reich is increasingly shared by others, his
specific work has nevertheless been criticized by trained economists, mostly on
the grounds that he is a lawyer, not a Ph.D. in economics. . . . That criticism
has been one reason that Mr. Reich . . . has failed to get a tenured professorship
at the Kennedy School, where he has been a lecturer . . . for more than a
decade. . . . The criticism of Mr. Reich's credentials . . . has intensified with
his appointment last week as chief of the economics transition team. (Uehitelle,
1992, p. 17)
But perhaps the most important reason for the persistent dominance of the
economic model as a way of understanding organizations is its implicit ideology.
The economic model presents a benign view of social organization. By stressing
markets and the operation of voluntary exchange, power, coercion, and exploita-
tion are left out of view. Employment relationships are voluntary—if someone
doesn't care for his or her employer, the person simply finds another one. One's
persistence in a particular exchange, including the exchange of work for wages,
is taken as prima facie evidence of the person's revealed preferences. The implica-
tion is that the particular exchange is preferred to other available alternatives, or
else the person would pursue one of those options. Thus, the economic model
permits and, in fact, encourages one to avoid dealing with or even thinking about
the potential conflict of interest between individuals and organizations and about
issues such as power and social control. The idea of voluntary exchange implies
that social control is accomplished readily and without coercion— and surely this
is a much more benign portrayal than in the ethnographic studies of actual work-
place relations (e.g., Burawoy, 1979; Graham, 1995; Kunda, 1992).
Granovettcr has also noted that the attractiveness of economic logic lies in
its way of dealing with the problem of social order, coordination, and control:
It has long been recognized that the idealized markets of perfect competition
have survived intellectual attack in part because self-regulating economic struc-
Five Models of Behavior 55
In this regard, economics shares with many other social sciences, including
some versions of organization theory, a perspective that sees "all that is real as
necessary, all that exists as inevitable, and thus the present mode of production as
eternal" (Braverman, 1974, p. 16). At the end, however, regardless of its ideologi-
cal comfort and the active exclusion of dissenters, there is little evidence that the
economic model's dominance in the literature derives from the ability to make
nontrivial predictions about organizational issues.
Economic models exclude social context or social relations almost entirely, except
tangentially and indirectly. For instance, one could potentially incorporate the
welfare of others or one's relationship with others into the typical economic utility
function, but the analyses seldom do so. Lazear (1995) did discuss peer pressure
in attempting to develop an economic rationale for team-based compensation.
But the inclusion of these more social considerations is still within the context of
individual utility maximization and choice: "Norms can be thought of more rigor-
ously as the equilibrium level of effort that results when an organization punishes
deviance" (pp. 49-50).
What might be termed a social model of behavior, in contrast, emphasizes
the embedded nature of behavior (Granovetter, 1985). P. M. Blau has noted,
"The fundamental fact of social life is precisely that it is social—-that human
beings do not live in isolation but associate with other human beings. . . . The
study of social structure . . . centers attention on the distribution of people
among different positions and their social associations" (1977, p. 1). Social models
emphasize that "one's behavior is rarely explicable without reference to previous
and persisting effects of interaction with others and the overall pattern of such
interactions in groups" (Granovetter, 1986, p. 31). Organizational behavior is
"embedded in concrete, ongoing systems of social relations" (Granovetter, 1985,
p. 487). Pfeffer and Salancik, outlining the resource dependence perspective on
interorganizational behavior, argued that "to understand the behavior of an orga-
nization, you must understand the context of that behavior. . . . Organizations
are inescapably bound up with the conditions of their environment" (1978, p. 1).
Social models of behavior emphasize the context of behavior more generally and
networks and social actors' positions in them and their social relations more spe-
cifically as causal explanations.
Several studies have demonstrated that individuals' perceptions of their work
environment, including task characteristics, are affected by what others say. White
and Mitchell (1979), O'Reilly and Caldwell (1979), and Weiss and Shaw (1979)
56 New Directions for Organization Theory
rector ties "render events that arc otherwise distant more proximate . . . creating
an immediate concrete example that may encourage imitation" (1993, p. 568).
Studying the adoption of the poison pill antitakeover defense, G. F. Davis
(1991) found that firms that had board of director interlocks to other firms that
had adopted poison pills were themselves more likely to do so, controlling for
other factors. "Being tied to one other adopter increased the rate by about 61
percent, ties to two other adopters nearly doubled the rate, and firms with six tics
to other adopters subsequently adopted at about 2.5 times the rate of comparable
firms with no ties" (p. 605). Mizruchi (1992) observed that interfirm linkages
affected the likelihood that firms would make campaign contributions to particu-
lar political candidates.
Strategy is also affected by interfirm ties. Greve (1995), studying the abandon-
ment of corporate strategy—in this instance, a particular radio format—observed
a social contagion effect from both market contacts and other units within the
same corporation. In a subsequent study concerning the adoption of a particular
radio format, Grcve (1996) observed a social contagion effect on the adoption of
a specific market strategy, with the contagion effect being particularly pronounced
for other stations with the same corporate ownership. Havernan (1993) studied
the entry of California savings and loans into six new markets over the 1977—1987
period: direct investment in real estate, mortgage-backed securities, nonresidential
mortgages, consumer loans other than mortgages, commercial loans, and other
services (e.g., insurance). For direct investment in real estate, nonresidential mort-
gages, and mortgage-backed securities, Haveman found evidence of imitation —
firms were more likely to enter a market in a given period the greater the number
of large firms and the greater the number of profitable firms that had previously
done so. Although this study did not measure interfirm linkages as precisely as
Davis or Greve, it provides additional evidence consistent with the idea that firms
are influenced by the behavior of others in their environment.
There is evidence that interfirm relationships help to account for organiza-
tional structures and do so net of technical factors. Palmer, Jennings, and Zhou
(1993) reported that interlocks with multidivisional firms influenced the adoption
of the multidivisional structure. L. R. Burns and Wholey (1993) studied the adop-
tion and abandonment of the matrix organizational form using a sample of more
than 1,300 hospitals over the seventeen-year period 1961-1978. Matrix manage-
ment is a particularly interesting structural innovation to examine, because it
"constitutes a shift from vertical-functional authority toward a hybrid, function-by-
product organization. . . . Matrix structures are team-oriented arrangements that
promote coordinated, multidisciplinary activity across functional areas, broad par-
ticipation in decisions, and the sharing of knowledge" (p. 107). As such, they are
almost diametrically different from the emphasis on hierarchy in transaction cost
economics. The empirical results showed no effect of organizational scale on
matrix adoption, but there were some positive effects of task diversity, a technical
factor predicted to be associated with the need for a matrix structure. Their results
show much stronger and more consistent effects of social influence on both adop-
tion and abandonment of matrix structures:
58 New Directions for Organization Theory
the prestige of a hospital influences not only its decision to adopt . . . but also
the decisions of neighboring hospitals. . . . Other significant effects suggest that
professional media . . . and regional and local hospital networks . . . are influ-
ential. Our results thus suggest that organizational networks influence the diffu-
sion of administrative innovations in much the same way that they influence the
spread of technological innovations, (p. 130)
Mizruchi and Stearns (1994) found that having a banker on a firm's board
increased the level of borrowing. Galaskiewic/ and Burt (1991) reported that cor-
porate officers responsible for philanthropy that were similarly positioned in the
interorganizational network evaluated potential contribution recipients similarly.
Galaskiewicz and Wasserman (1989), studying patterns of corporate giving in the
Minneapolis-St. Paul area, found numerous examples of the importance of net-
work connections. Nonprofits received more money from corporations that pre-
viously gave money to other nonprofits whose directors sat on the focal nonprofit's
board. And, firms gave more money to nonprofits that were previously funded by
companies whose CEOs or philanthropy officers were personally known by the
firm.
Haunschild (1993), examining the acquisition activity of 327 firms, found
that firms imitated the acquisition activities of firms to which they were tied
through directorships. In a subsequent study, Haunschild (1994) found that the
size of the acquisition premia paid by firms in takeovers was related to the premia
paid by their interlock partners as well as to the size of premia paid by other
acquirers using the same investment bank. The evidence from the studies of char-
itable giving and corporate takeover activity is consistent with the idea that net-
work influences are particularly important when there is more uncertainty, for
social communication is useful in resolving that uncertainty (Fcstingcr, 1954).
Network Concepts
There arc a number of important concepts and ideas in network analyses of orga-
nizations. One concept has to do with the fact that, among social actors, "weak
ties" often offer advantages over stronger ties. One of the ironies uncovered by
Granovetter (1973; 1974) was the relative importance and effectiveness of com-
paratively weak ties among individuals in the job-finding process. Although one
might think that strong ties, such as friendship or kinship, would be more useful
in activities such as job finding, this was not the case. Although strong ties create
a closer bond between individuals, people to whom one is more strongly tied are
more likely to be in the same social network and thus provide redundant contacts
and information. Weak ties link individuals to different and more diverse sets of
contacts and consequently provide less-redundant information and social relations
compared to strong ties.
Network models of organizations have been developed most completely by
Burt (1992), who has also developed a computer program, Structure, to imple-
ment his method of social network analysis. Hurt's basic ideas of structure concern
structural equivalence, constraint, and structural holes. If a social unit has the
same relations with others as another unit (which could be an individual or a
firm), the two social entities are considered to be structurally equivalent. Because
they have the same pattern of social relations, the entities are competitors, even
if they arc not aware of each other. Burt has noted, "competition is a matter of
relations, not player attributes" (p. 3). Reanalyzing data on the diffusion of the
antibiotic tetracycline (Coleman, Katz, and Menzel, 1966), Burt (1987) found
Five Models of Behavior 61
that structural equivalence was a more powerful predictor of diffusion than was
social contagion. In social contagion models of diffusion, the argument is that
information or an innovation diffuses through direct social contact or ties (e.g.,
you talk to a colleague who tells you about the new product, and then you are
more likely to adopt it). This mode of analysis, emphasizing direct social contact,
has a long history in sociology. For instance, Lazersfeld, Berelson, and Gaudet
(1944) demonstrated that individuals who voted for a particular political party
tended to have friends affiliated with that party. Hurt found that individuals in
structurally equivalent positions— competitors—were more likely to adopt some
innovation once it had been adopted by their competitors even in the absence of
direct contact. Thus, competition, as defined by the network measure of struc-
tural equivalence, was a better predictor of diffusion than social contagion.
Constraint is also defined in terms of one's network of relations. A given
organization, for instance, is constrained to the extent that it transacts with other
organizations and has few alternatives to those organizations, while the organiza-
tions themselves have diffuse or unconcentratecl transactions. Conversely, one has
power in exchange relations to the extent that one has options and transacts with
others who themselves are highly constrained. Burt (1980; 1983) has shown that
measures of network constraint can predict profitability and interorganizational
relationships, such as board of director interlocks, designed to co-opt or otherwise
mitigate such constraints.
Finally, a structural hole represents separation between nonredundant con-
tacts and is similar, in some respects, to the idea of weak ties. When contacts,
members of one's own network, are not themselves linked, Burt (1992) argued
that there could be profit in serving as an intermediary—not directly linking the
parties but mediating transactions among them. Burt has theoretically derived and
empirically tested the idea that "players with a network optimized for structural
holes, in addition to being exposed to more rewarding opportunities, arc also
more likely to secure favorable terms in the opportunities they choose to pursue"
(p. 30). This is because the social actor in the middle can play the parties off
against each other and can also profit from arranging transactions that otherwise
might not occur. Indeed, Burt has defined this role of being an intermediary as
an important component of entreprencurship.
Recently, Podolny (1993) has developed a fourth important concept for un-
derstanding social networks, that of status. Status is defined as "the perceived
quality of that producer's products in relation to the perceived quality of that
producer's competitors' products" (p. 830). Status is determined not only by the
objective quality of the good or service but, particularly when that quality is diffi-
cult to directly assess, by the ties a given organization maintains with buyers,
other producers, and third parties. For a producer of a given level of quality, status
translates into higher revenues and lower production costs (p. 837). Civen the
advantages of attaining higher status and the advantages that accrue to those
organizations with high status, two important questions emerge. First, what keeps
other organizations from acquiring status and therefore disturbing the stability of
the status order? And second, why doesn't one or a subset of the highest-status
producers come to totally dominate the market?
62 New Directions for Organization Theory
The answers to both questions come from the fact that status is produced, in
part, through network linkages, and this fact provides both opportunity and con-
straint. Although lower-status organizations may seek to form linkages with higher-
status producers in order to enhance their status, there is little or no incentive for
the higher-status organizations to want those linkages formed. Indeed, to the ex-
tent that higher-status organizations do form linkages with lower-status entities,
their own status is potentially at risk. This idea also helps to answer the question
about the limits of market dominance:
Since the relationship between actual and perceived quality is mediated by the
producer's tics to others in the market, the producer invariably changes how it is
perceived if it broadens relations with others in the market. . . . Ib the extent
that a higher-status producer attempts to expand into the position of a lower-
status competitor, it changes its reputation and thus alters the cost-and-revenue
profile that provided it with the initial advantage. As a result, just as status pro-
cesses help reproduce inequality by constraining those at the bottom of the status
hierarchy, so status processes also place limits on the higher-status producer's
expansion into the lower end of the market. (Podolny, 1993, pp. 844-845)
Podolny (1993) showed that the percentage spread in debt underwritings (the
gross spread divided by the dollar amount of the bond offering) was negatively
related to the status of the underwriting firm, indicating that, as predicted, higher-
status investment banks enjoyed a cost advantage that resulted from their status
and permitted them to underbid their lower-status competitors. Podolny (1994),
studying which banks formed relationships with others, found that there was an
effect of market uncertainty on this process. When uncertainty was greater, orga-
nizations tended to engage in transactions with those with whom they had dealt
in the past, and they tended to engage in transactions with those of similar status.
These results suggest that the status hierarchy will be particularly stable under
conditions of greater uncertainty, because it is under precisely those circum-
stances that transactions, and presumably other relationships, will be most re-
stricted by history and status similarity.
In some sense, network models of organizations arose out of the resource depen-
dence perspective (Pfeffer and Salancik, 1978), which was filled with network
imagery and concepts but did not employ formalized network measures or meth-
ods in the empirical research. Both approaches depart from the "rationalist, indi-
vidualistic perspective" that "has denied both the reality of organizations as insti-
tutions . . . as well as the embedded and at times quasi-political character of
organizational action and choice" (Pfeffer, 1987, p. 25). The resource depen-
dence perspective, however, focuses attention not just on the social effects on
behavior but also on the fact thai organizations have a reality apart from "serving
as settings and apart from the various people who were in them at any point in
time" (p. 26). As such, the theory is a reaction to the view, increasingly common
in both economics and political science, that "formally organized social institu-
Five Models of Behavior 63
(1) the fundamental units for understanding intereorporate relations and society
are organizations; ours is a society of organizations (Presthus, 1978); (2) these
organizations are not autonomous, but rather are constrained by a network of
interdependencies with other organizations; (3) interdependence, when coupled
with uncertainty about what the actions will be of those with which the organiza-
tion is interdependent, leads to a situation in which survival and continued suc-
cess are uncertain; and, therefore, (4) organizations take actions to manage exter-
nal interdependencies, although such actions are inevitably never completely
successful and produce new patterns of dependence and interdependence. Fur-
thermore, (5) these patterns of dependence produce interorganizational as well
as intraorganizational power, where such power has some effect on organizational
behavior. Organizations tend to comply with the demands of those interests in
their environment which have relatively more power. (Pfeffer, 1987, pp. 26-27)
Institutional theory is another important variant of a social model of behavior.
Scott argued that institutional theory broadened the conception of the environ-
ment considered in some other organization theories: "To the earlier emphasis
on the importance of the technical environment—resources and technical know-
how—institutional theory has called attention to the importance of the social and
cultural environment, in particular, to social knowledge and cultural rule sys-
tems" (Scott, 1995, p. xiv). One variant of institutional theory has emphasized
cognitive, mimetic processes in which organizations, particularly when confronted
with uncertainty, resolve that uncertainty by imitating what appears to be preva-
lent practice (DiMaggio and Powell, 1983). Scott wrote: "We seek to behave in
conventional ways . . . that will not cause us to stand out or be noticed as differ-
ent. . . . We attempt to imitate others whom we regard as superior or more
successful. One principal indicator of the strength of such mimetic processes is
prevalence: the number of similar individuals or organizations exhibiting a given
form or practice" (p. 45). Another variant of institutional theory emphasizes con-
formity deriving from the taken-for-granted or assumed character of social life.
Zucker noted that "institutionalization is rooted in conformity—not conformity
engendered by sanctions . . . nor conformity resulting from a 'black-box' internal-
ization process, but conformity rooted in the takcn-for-granted aspects of everyday
life. . . . Institutionalization operates to produce common understandings about
what is appropriate and, fundamentally, meaningful behavior" (1983, p. 5).
64 New Directions for Organization Theory
Although some (e.g., DiMaggio, 1988) have argued that institutional theory
deemphasizcs the role of interests as contrasted with resource dependence or
network perspectives, Scott argued that "institutional theory reminds us that inter-
ests are institutionally defined and shaped" (1987, p. 508). What institutional
theory has emphasized is the role of the state and external authority in imposing
organizational arrangements and the fact that organizational structures and pro-
cesses may serve to legitimate organizations rather than to advance technical effi-
ciency.
In comparison to economic models, social models of behavior dcernphasize
individual characteristics but rather emphasize the relationship and connection
among social actors. One achieves promotions, for instance, not so much as a
consequence of one's individual human capital (Becker, 1964) but because of
one's location in the social structure. Turnover is related to the turnover of one's
social contacts (Krackhardt and Porter, 1986), job attitudes are affected by the
attitudes of others in one's social environment (e.g., Thomas and Griffin, 1983)
and particularly by those in one's communication network (Ibarra and Andrews,
1993), and one's ability to find a job depends importantly on social relations
(Granovetter, 1986). Status and associated perceptions of quality depend on the
status of those organizations (or individuals) to which one is tied (Podolny, 1993).
Organizational survival and economic success depend on the pattern of connec-
tions to other organizations (Pfcffer and Salancik, 1978).
Although we know that organizations are, fundamentally, relational entities,
and that the environment of organizations is composed of other organizations and
filled with transactions and other interchanges among them, many theories and
analyses take the individual or the organization alone as the unit of analysis
and fail to incorporate ideas or measures of social structure into either theory or
empirical research (Pfeffer, 1991, p. 800). The social model of behavior avoids
this pitfall even as it challenges more individualistic conceptualizations of organi-
zational and intcrorganizational behavior.
ceived interlocks compared to neutral interlocks, which are often called weak
when compared to more direct interlocking. Some studies find effects of neutral
interlocks (Palmer et al., 1995) and not of stronger ties, but there arc few a priori
predictions about which type of tie should be important in a given study. Re-
viewing much of this literature, Fligstein and Brantlcy concluded that "we should
abandon our concentration on ... directors as a source of network data, in gen-
eral, and financial linkages across boards of directors, in particular, unless their
possible relevance can be specified theoretically" (1992, p. 304).
Further complicating the issue, many studies that ostensibly use network lan-
guage don't employ network methods for analyzing the actual structure of the
network but use very reduced measures instead. For instance, Palmer ct al. (1995)
measured connections in terms of the number of sent industrial and commercial
bank interlocks, the number of received industrial and commercial bank inter-
locks, and the number of neutral industrial and commercial bank interlocks.
They measured industrial and banking interlocks separately because of the im-
portant role of banks in some theories of how organizations are linked (e.g., Soref
and Zeitlin, 1987). Numbers do not indicate the importance of the particular
organizations linked by a tie, and these studies typically don't measure centrality,
closeness or proximity, and similar indicators of network structure.
Using formal network models requires a great deal of data on relations among
individuals or organizations, data which often are not available. Thus, for in-
stance, studies of interorganizational relations from a resource dependence or
structural autonomy perspective use data not on the specific transaction interde-
pendencies of given organizations but rather the patterns of transactions among
economic sectors. This is because specific organizational transactions data are not
available. As Finkelstein (1996) noted, however, this presents a problem of aggre-
gation bias in the data. Gathering data on ties among individuals to model career
mobility (Burt, 1992) requires a lot of information, so such studies are typically
confined to just one or a very few organizations. But this type of design necessarily
limits the researcher's ability to explore conditions under which network effects
are stronger or weaker or to compare measurement models directly in different
settings. In this sense, network models of organizations are richly detailed depic-
tions of social relations, but the empirical specificity constrains the empirical re-
search process.
Finally, it should be acknowledged that the social model of organizational
behavior is a useful and productive way of understanding how information, struc-
tures, decisions, and other policies and practices diffuse across social actors. This
model, however, does not have as much to say about how the various practices or
structures arise in the first place. In that sense, the social model is potentially a
useful complement to other perspectives on the source of behavior in organiza-
tions.
(Aronson, 1972). The retrospectively rational model suggests that individuals and
organizations will take actions to make sense of or to appear to be consistent with
previous choices. There are two important variants of the retrospectively rational
model —one that emphasizes a "cooler" cognitive process of self-perception and
one that implicates "wanner" social psychological processes such as commitment
and self-justification.
The basic premises of the self-perception model are well enough known in
psychology to bear only brief recitation. First, behavior may be initially mindless,
habitual, or automatic (Langer, 1983) or, alternatively, produced by other salient
influences; behavior is not always prospectively chosen to be consistent with pref-
erences—indeed, one of the intended consequences of engaging in some activity
is discovering what one's true preferences are (J. G. March, 1978). Second, indi-
viduals, particularly when questioned or called to account for their behavior, pre-
fer to appear sensible, logical, and consistent, if for no other reason than as part
of a strategy of self-presentation (Alexander and Knight, 1971); in this sense, indi-
viduals act as if they were lay social scientists inferring the causality of their own
behavior (Kelley, 1971). Third, when questioned about their attitudes or behav-
iors, individuals will look to their own past statements and actions and to the
external environment for explanations as to the sources of those attitudes or be-
havior; this process has been referred to as self-perception (Bern, 1972) or retro-
spective rationality (Staw, 1980). Because there arc social norms valuing consis-
tency, individuals will often attempt to behave in a fashion consistent with the
attitudes and beliefs they have inferred they hold by retrospectively examining
their own past behavior, thereby creating a situation in which the past comes to
affect the future through its influence on self-perceptions and attitudes.
Weick's (1964) study of the effect of cognitive dissonance on performance is
a good example of studies of this type. In the study, individuals agreed to do a
difficult cognitive task after they learned they would get a smaller reward for
participating in the study than they had expected. "Those that were deprived most
severely rated the subsequent task more interesting and were three times more
productive at performing it than were people who felt less deprived" (Weick,
1993, p. 11). Faced with the prospect of a greatly diminished reward for some-
thing they were committed to do, individuals made sense out of their commit-
ment by evaluating the intrinsic interest and worth of the task more highly. And,
"by working hard and solving more problems, they created a situation that con-
firmed the sensibleness of agreeing to participate in the first place" (p. 11). Thus,
a process of commitment caused individuals to cognitively revalue a task, and
having engaged in that revaluation, they proceeded to act differently, working
harder than those that had not gone through the same degree of cognitive revalu-
ation created by the diminished external reward.
Thus, one particularly important formulation of this perspective notes that if
there are no salient extrinsic reasons for engaging in some activity—such as large
external rewards or the threat of important sanctions —individuals will come to
believe that they are doing the activity because they truly like it. This is the
insufficient justification paradigm that is discussed thoroughly in the literature on
behavioral commitment (Kiesler, 1971; Salancik, 1977) and cognitive dissonance
Five Models of Behavior 67
more committing program. But those men who were committed to ROTC but
discovered they had a high draft number (low probability of being drafted) faced
a problem. They joined ROTC to avoid the draft and serve under better circum-
stances as commissioned officers, but now they learned that they might not have
had to serve; they no longer had an extrinsic reason for being in ROTC. But
because of their contractual commitment, they could not withdraw. As might be
expected under insufficient justification predictions, this group earned higher
grades in their ROTC courses and expressed the most favorable attitudes toward
the program compared to the cadets with extrinsic reasons for being in ROTC.
They made sense of their commitment in the presence of insufficient external
justification by finding, within the program, benefits and elements to enjoy and
appreciate.
O'Reilly and Caldwell (1980), in a study of 108 MBA graduates, examined
the effects of the bases of job choice on job satisfaction and commitment. They
found a "positive relationship between intrinsic decision factors [intrinsic interest
in the job itself, personal feelings about the job, etc.] and satisfaction and commit-
ment" (p. 561). They also found evidence that "decisions made under external
constraints, that is, decisions based on a concern for family and financial consider-
ations, are inversely related to job satisfaction and commitment" (p. 563). O'Reilly
and Caldwell (1981) studied the effects on MBA graduates of the number of
offers they received and whether or not they took the highest offer. They predicted
that graduates who had several job offers and who accepted a job that did not pay
the best among the offers should have fewer extrinsic reasons for their choice
than those who either took the best offer or had only one offer and, consequently,
should be more committed to the job. They found that, consistent with this pre-
diction, individuals who took their initial offer with fewer external reasons for
doing so in fact were more satisfied with the job and stayed in it longer.
In a third field test of these ideas, Caldwell, O'Reilly, and Morris (1983)
studied both satisfaction with and intrinsic interest in an evening MBA program
on the part of 132 students, some of whom received more generous employer-
provided subsidies for attending the program than others. They found that suf-
ficiency of justification effects depended both on orientation toward the activity
and beliefs about the subsidy. Specifically, "providing an educational subsidy
is associated with lowered task interest for employees who are expressively ori-
ented. . . . Failure to provide an extrinsic reward is related . . . to increased
interest for those who do not [believe] the subsidy is a normal employee benefit"
(p. 506).
One implication of this retrospective rationalization argument is that once
individuals are committed to some organization or activity, additional incentives
probably will not increase their motivation or positive affect and can, in fact,
undermine their positive attitudes and actions. Pfeffer and Lawler (1980), studying
college and university faculty, found a positive relationship between overly suffi-
cient justification (salary was higher than would be predicted based on experi-
ence, productivity, and other characteristics) and job satisfaction for faculty who
had been in their organizations for a shorter period of time. For faculty with
longer organizational tenure, who were more committed to the organization as a
Five Models of Behavior 69
Staw and Hoang (1995) demonstrated the effect of commitment and sunk
costs in a field study of the National Basketball Association. They showed that
draft order—how early a player was drafted—was a statistically significant pre-
dictor of the amount of playing time even over a five-year period, controlling for
performance in the preceding year, position, and injuries. They also found that
draft order affected a player's career length and the likelihood of being traded,
again with performance and other factors statistically controlled. Players taken
earlier in the draft were less likely to be cut from a team or traded. Staw and
Hoang discovered that being traded was negatively associated with subsequently
being retained by the new team and with playing time, even with performance
held constant. They interpreted this latter result to mean that "the further teams
were from the original drafting of players, the less committed they were to them"
(p. 491), and this reduced commitment had predictable effects on player use and
retention.
Another arena in which commitment effects have been demonstrated is the
performance appraisal process. Bazerman, Beekun, and Schoorman (1982), in an
experimental simulation of a managerial situation that involved making a promo-
tion decision, found that "subjects who were responsible for the promotion deci-
sion gave higher pay increases and bonuses, evaluated the manager's potential
more favorably, and projected higher sales and earnings than did the subjects
who were not responsible for the promotion decision" (Schoorman, 1988, p. 58).
Schoorman (1988) replicated this result in field research conducted in a large
public sector organization. He found that when the individuals filling out perfor-
mance appraisal ratings had input into and agreed with the promotion or hiring
decision for the person being rated, the performance appraisal score was more
positive than when the rater did not have input into the decision. Schoorman
reported that "the escalation effects accounted for 6% of the variance in perfor-
mance appraisal ratings" (p. 60), which was more than the variance accounted
for by the measures of clerical ability that were used to actually select employees
for the jobs. Having once made or agreed with a hiring or promotion decision,
the way to make sense of or rationalize that choice was apparently to evaluate the
performance of the selected individual more positively.
suggest that people escalate commitment when they do not set a budget and
when incremental costs and benefits are difficult to track (p. 47). Heath summa-
rized the existing literature:
There is currently no evidence that people will make errors of escalating commit-
ment when marginal investments are explicit. . . . The literature provides no
clear demonstration of such errors because previous demonstrations of "escalat-
ing commitment" with explicit investments did not provide relevant information
about marginal costs and benefits. . . . When researchers omit information
about benefits . . . their studies may not control subjects' expectations about the
benefits of a marginal investment. Subjects may assume that benefits increase as
sunk costs increase. . . . Other studies . . . do not control for expectations about
costs. (1995, p. 52)
Heath (1995), in a series of experiments, tested the idea that people set men-
tal budgets for investments and resist exceeding those budgets as they track their
expenditures against the budgets. He warned about the problem of de-escalation
of commitment even in the presence of high marginal returns because of adher-
ence to the mental budget model:
when benefits come from completing a project, investments toward the end of a
project produce increasingly high marginal returns. In situations where there is
learning, previous investments may decrease marginal costs and increase mar-
ginal benefits by moving people clown a learning curve. Mental budgeting pre-
dicts that in such situations people are more likely to make errors of de-escalating
commitment (p. 53).
being placed on consistency and persistence: individuals who persist can come to
be evaluated more positively than those who don't maintain their commitments,
making perseverance rational for those individuals.
Commitment can also be sensible in that the very commitment can have
positive effects on results. Consider, for instance, the Staw and Hoang (1995)
study of the effects of draft order on playing time and being retained in the NBA.
Why might it make sense to play someone more than might be warranted by that
individual's performance in the previous year? Because there is evidence that
playing time itself positively affects future performance: "We regressed perfor-
mance for years 2 through 5 on prior year's performance, prior year's time on the
court, draft number. . . . The results of these four analyses supported a straight-
forward investment hypothesis. . . . An increase in performance was associated
with the investment of prior playing time" (p. 489).
Moreover, Staw and Hoang wrote about "the vagaries of forecasting talent"
(p. 477) and also did an analysis that demonstrated that "draft number" was "a
significant predictor of subsequent performance during players' second and third
years in the NBA" (p. 488). Consequently, "draft order does appear to contain
some useful information." Given these facts, some persistence in following one's
initial predictions concerning talent might appear to be justified.
In the world outside of the experimental laboratory, the rationalizing model
may be more rational than sometimes thought. That is because performance is
not necessarily fully determined by preset factors and is affected by expectations
and by the resources that individuals receive as a consequence of commitment,
such as an opportunity to gain more experience. Moreover, performance is often
ambiguous in terms of its assessment, which means that acts of commitment can
change the evaluation of results. In this real world, the effects of both self-
perception and commitment on subsequent behavior and performance help to
create the very results that the rationalizing model seek to explain. These feed-
back cycles —resulting from both self-perception and commitment—to subse-
quent attitudes, behavior, and dimensions of performance would seem to warrant
more attention not only to evaluate the ability of the rationali/ing model to ac-
count for behavior but also to establish its implications for actually managing
behavior.
Recently, there has been growing interest in models of behavior that move away
from the neoclassical economic paradigm but in some instances retain a focus on
individual or organizational rational choice. These models differ from the eco-
nomic model of behavior in their core assumptions. Etzioni (1988, p. 4) has
articulated the core assumptions of one behavior model that includes morals.
First, a moral model postulates that individuals pursue not only pleasure (as they
do in the utilitarian approach) but also morality. Consequently, "individuals are,
simultaneously, under the influence of two major sets of factors — their pleasure
and their moral duty" (p. 63). Second, this perspective emphasizes the fact that
74 New Directions for Organization Theory
individuals choose means, not just goals, and these means are chosen on the basis
of their values and emotions. Because moral and utilitarian considerations are
qualitatively different, they cannot be traded off against each other (p. 71). An
individual obtains "a sense of affirmation . . . when a person abides by his or her
moral commitments" (p. 36). Third, the neoclassical presumption of the primacy
of the individual is modified to account for the fact that social collectivities, in-
cluding organizations but also including neighborhood and ethnic communities,
arc important decision-making units.
Frank (1990) has counterposcd what he has called the self-interest model
with a commitment model of behavior. Frank (1988; 1990) first reviewed the
various "rational" reasons to forgo short-term self-interest, such as building a repu-
tation for being trustworthy, as well as evolutionary or selection mechanisms that
might produce more cooperative behavior. For example, the fact that people who
cheat in transactions and are discovered will probably not have too many more
opportunities to cheat again means that those who remain in the arena of ex-
change are probably not going to behave completely opportunistically. Frank dis-
tinguished these more sophisticated versions of self-interest from his view of com-
mitment, which he described thus: "The motive is not to avoid the possibility of
being caught, but to maintain and strengthen the predisposition to behave hon-
estly. . . . It is this change in my emotional makeup . . . that other people may
apprehend" (1988, p. 95, emphasis in original).
There is extensive empirical evidence consistent with the moral model of
behavior, particularly as it is contrasted with a model of attitudes and behaviors
resulting from the operation of simple self-interest. However, most of the relevant
research does not deal with organizational contexts or problems, which leaves an
important research agenda — applying the ideas of the moral model of behavior
to organizations.
There have been studies showing that political behavior is not completely
explainable by the idea of self-interest. Voting is a public-spirited act—it requires
some effort, but the odds that a given individual's vote will be consequential for
determining the election outcome is almost infinitesimally small. Flirschleifer
(1985) reviewed studies showing that there was a correlation between the ex-
pected closeness of the election and voting—people were more likely to vote
when the election was thought to be close, as self-interest would predict—but the
relationship was comparatively small. Barry (1978) presented data indicating that
closeness of the election accounted for about 9 percent of the variation in voter
turnout, while a citizen duty score, a measure of moral obligation, accounted for
between 30 percent and 40 percent.
Scars and Funk (1990) reviewed numerous studies of the role of self-interest
in accounting for the political opinions of Americans on issues including race,
taxes, crime, and war. They concluded that only about one-quarter of the self-
interest terms in regression equations met even minimal standards of statistical
significance and that the average bivariate correlation between a measure of self-
interest and the corresponding political attitude was about 0.07, indicating that
even when statistically significant, self-interest contributed in only a small way to
explaining political attitudes. By contrast, symbolic predispositions, which include
Five Models of Behavior 75
The logic of the problem of collective action is both compelling and illuminat-
ing. The problem with this logic, however, is that the empirical evidence for its
ability to predict actual behavior is either scant or nonexistent. . . . Under the
conditions described by the theory as leading to free riding, people often cooper-
ate instead. . . . In over 13 experiments we have found that subjects persist in
investing substantial proportions of their resources in public goods despite condi-
tions specifically designed to maximi/c the impact of free riding and thus mini-
mize investment. The prevalence of such economically "illogical" behavior was
replicated over and over again. (Marwell, 1982, pp. 208, 210)
Etzioni,(1988) has reviewed a set of studies that indicate that for moral issues,
activities or behavior that have been deemed to be sacred, the calculative ap-
76 New Directions for Organization Theory
proach does not do very well in explaining behavior and, instead, nonrnarkets
exist. He concluded, "in any given society and period, certain items are not con-
sidered legitimate material for exchange. . . . As a result, market models are
expected to depict the relations poorly at best; socio-economic conceptions that
include moral as well as economic factors are necessary" (p. 81).
Institutional theory in some of its variants has a normative component: "em-
phasis . . . is placed on normative rules that introduce a prescriptive, evaluative,
and obligatory dimension to social life" (Scott, 1995, p. 37). Sclznick defined
institutionalization as "to infuse with value" (1957, p. 17). Holm has noted that
"the study of institutions is the study of norm-governed behavior" (1995, p. 398).
J. G. March and Olsen, arguing for the importance of institutions for understand-
ing politics, noted the importance of social obligations in understanding behavior:
"To describe behavior as driven by rules is to see action as a matching of a
situation to the demands of a position. Rules define relationships among roles in
terms of what an incumbent of one role owes to incumbents of other roles" (1989,
p. 23, emphasis added).
An interesting illustration of the effect of the normative order on the adoption
of specific organizational practices is provided by 'Iblbert and Zucker's (1983)
study of the diffusion of municipal civil service reform in the United States during
the period 1885-1935. Early in the period, which cities adopted reforms could
be predicted on the basis of characteristics that might be reasonably argued to
account for the need and desire for reform: larger cities, cities with a larger pro-
portion of immigrants, and those with a higher proportion of white-collar to blue-
collar workers were more likely to adopt reform. Although city characteristics
were strongly predictive of adoption of civil service reforms during the early pe-
riod from 1885 to 1904, the effects of these characteristics became weaker over
time so that by 1935, city characteristics no longer had predictive power. 'Iblbert
and Zucker interpreted this residt as demonstrating normative influence: "As an
increasing number of organizations adopt a program or policy, it becomes progres-
sively institutionalized, or widely understood to be a necessary component of ra-
tionalized organizational structure. The legitimacy of the procedures themselves
serves as the impetus for the later adopters" (p. 35).
The normative order is importantly influenced by the actions of the state,
which has important coercive and regulatory powers. Baron, Dobbin, and Jen-
nings (1986), examining the diffusion and implementation of modern personnel
practices, observed an important discontinuity at the time of World War TI. At
that point, the state intervened in organizational administration — for instance,
forbidding strikes, demanding personnel records so that an inventory of manpower
could be maintained and managed, and limiting wage increases. Baron et al.
noted that "these interventions . . . fueled the development of bureaucratic con-
trols by creating models of employment and incentives to formalize and expand
personnel functions" (p. 369). Bridges and Villemez (1994), analyzing a matched
sample of employees and employers from Chicago in the early 1980s, found that
both government influence in the industrial sector and how close organizations
were to the public sector had important effects on the development and spread
of due-process procedures and bureaucratic control.
Five Models of Behavior jj
The conclusion from this brief review of relevant empirical evidence seems
inescapable: social and individual values matter in understanding behavior and
are often more powerful predictors than simple conceptions of self-interest. In
determining the normative order, both prevalence and the actions of legitimate
authorities, such as governments, are important. Consequently, understanding the
emergence of norms and moral conceptions in organizations, and examining their
effects on behavior, are important research tasks.
Conclusion
loosely related to the degree of empirical support they enjoy or the reasonableness
of their assumptions. Apparently, models of behavior often take on a religious
quality, adopted or rejected on the basis of belief or aesthetics rather than on the
basis of scientific evidence. So, even though organizations are inherently social
and relational entities, there is great interest in the economic model of behavior
in spite of the fact that many of its variants proceed from a position of method-
ological individualism that denies the very reality of the institutions and organiza-
tions being explained. Even though self-perception and commitment processes
have been repeatedly demonstrated in both field and experimental situations, and
these perspectives have important implications for the effects of surveillance, in-
centives, and other forms of external control, the insights of these models are
seldom incorporated in either the research on or design of organizational incen-
tive systems. Even though moral considerations loom large and often dominate
self-interest in studies of political behavior, a moral model of organizational be-
havior has yet to be developed or explored to any great extent.
Perhaps the greatest difficulty is that the models often seem to simply ignore
each other. Empirical explorations that attempt to compare and contrast at least
some of the various competing models of behavior—the economic model, the
social model, the retrospectively rational model, and the moral model — are sub-
stantially less common than warranted. Because these theoretical perspectives af-
fect the practice of management, the formulation of public policy, and the eon-
duct of research, attention to their assumptions, evidentiary base, and implications
would seem to be critical for anyone doing research on organizations.
4
As discussed thus far in this book, underlying all organization theory and research
is the basic question, How are we to understand organizations? One variant of a
social model of behavior that has drawn sufficient research attention to merit
more extended consideration is an approach to organizational analysis emphasiz-
ing the effects of social composition.
Although many organization theories have psychological or social psychologi-
cal foundations (Mowday and Sutton, 1993), which emphasize attitudes, disposi-
tions, and cognitive processes, one important portion of the field employs a struc-
tural perspective. "This focus on structure tends to direct attention . . . toward
the enduring properties of the relations among actors that both constrain and
enable action to occur" (Pfeffer, 1991, p. 789). Structural analysis explains "pat-
terns of social relations in terms of properties of social structure, not in terms of
the assumptions made, whether or not these are derivable from psychological
principles" (P. M. Blau, 1977, p. 246). Although structural accounts may rely on
psychological principles or assumptions, these are not always directly tested or
even examined in the analyses. The structural perspective relies on the social
model of behavior for its core assumptions and so emphasizes the effect of context
on behavior. One advantage of analyzing organizations in terms of the effects of
their composition is that such analyses rely on more readily observable variables
(Pfeffer, 1983; Hambrick and Mason, 1984). Wcick has noted the advantages of
analyzing organizations in terms of observable behaviors and other properties:
There are several plaees in the organizational literature where investigators seem
to resist defining their concepts in terms of observable actions by individuals
in the mistaken belief that, in doing so, they will have to explain the actions
psychologically. If ... properties can be defined in terms of observable individ-
ual behaviors, there is a better chance the empirical research . . . can be made
more cumulative. (1969, pp. 31-32)
81
82 New Directions for Organization Theory
One subtle but important implication of the social model, with its emphasis
on the effects of social influences, is that not every social entity is going to be
equally available to exert influence. Thus, there is the argument that patterns of
social relations are explained not solely on the basis of motives but "on the basis
of external opportunities for social relations created by population composition
and the structure of positions in the social environment" (Iluber, 1990, p. 2,
emphasis added). What this means is simply that who you interact with and are
therefore influenced by and, in turn, influence, is a function of the existing social
structure. If you never have contact with Democrats because of the particular
composition of your social environment, your political attitudes will be affected
accordingly. If an individual's social environment is made up almost exclusively
of males, the opportunity for interaction with women is limited simply by the
composition of the social group, regardless of the individual's interest in inter-
acting with women. This latter example is particularly relevant, as we will see
below, in understanding women's networks and careers in numerically male-
dominated environments. As an empirical illustration of this compositional effect,
P. M. Blau and Schwartz (1984) explored patterns of intermarriage among racial,
occupational, ethnic background, and birth-region dimensions in a sample of
metropolitan areas. They acknowledged that individuals would prefer to associate
with similar others. However, exercising this preference is constrained by the
available interaction partners. Blau and Schwartz demonstrated that the relative
size of various demographic groups affected the observed rate of intermarriage.
Many explorations of structural effects have examined the impact of varying
group composition in terms of salient demographic differences such as age, orga-
nizational tenure, gender, or work background (Pfeffer, 1983; Stewman, 1988).
These studies usually proceed from the empirically justifiable presumption that
similarity is an important basis of interpersonal attraction and contact (Bcrscheid
and Walster, 1969) as well as the assumption that background and contact matter
for understanding behavior. In the organizations literature, structural effects have
been explored most completely in the literature on organizational demography, a
related literature on the effects of the composition of top management teams, and
a literature that has explored the effect of gender composition on work-related
outcomes such as wage and position attainment. A part of the literature on gender
composition examines how this affects social networks, and through such network
connections, mobility prospects.
Organizational Demography
Kanter's (1977) argument that proportions have an important effect on the experi-
ences of those in the minority was an explicit recognition of compositional effects.
She wrote, "If one sees nine X's and one O, the O will stand out" (p. 210). Kantcr
argued that women in token status suffered from being more visible and thus
under more intense scrutiny and came to be stereotyped, seen as representative
of the group (women) rather than as distinct individuals. Differences between
those in token status and the majority were exaggerated because the presence of
Effects of Organizational Composition 83
persons who were different: made the similarities among those in the majority
more salient. Kanter argued that the pressures that confronted someone in token
status made success difficult, particularly in organizational contexts such as man-
agement positions in which trust and interpersonal communication were im-
portant. Her research identified three sources of stress:
Because their "differentncss" is highly visible, tokens feel that they are always
under scrutiny. . . . Because they are symbolic representatives of their "type,"
tokens experience added pressure to perform well, since this may determine fu-
ture opportunities for other individuals in their social category. . . . One re-
sponse to performance pressure is to overachieve; another is to avoid calling
attention to oneself or one's accomplishments. . . .
The second source of stress, boundary heightening, results from majority
group members' tendencies to exaggerate their own commonalities as well as
their differences from tokens. . . .
The third source of stress, role entrapment, involves typecasting by domi-
nants. (P. B. Jackson, Thoits, and Taylor, 1995, p. 545)
geneity in the group as a whole and the amount of group-level social integration).
Their "results show group-level tenure homogeneity to be associated with lower
turnover rates of individuals, with group-level social integration as a moderator.
. . . Furthermore, this process is not merely the aggregate of an individual-level
process, since individual-level social integration does not affect the turnover rates"
(p. 33). The results of this study demonstrated substantive as well as statistical
significance. Individuals in the group with the least social integration had an
estimated turnover rate nine times as high as would be expected without this
effect of demographic diversity. By contrast, individuals in the most integrated
group had an estimated turnover rate 94 percent lower than it would have been
without this effect of demographic cohesion.
A further study of processes that undergird the effects of composition and
demography demonstrated that the greater the work-unit diversity, the lower the
levels of psychological attachment, operationalized as commitment, attendance at
work, and turnover intentions (Tsui, Egan, and O'Reilly, 1992). This study also
found that the effects of heterogeneity were not constant across demographic
groups, with whites and males evidencing larger negative effects from being in
more demographically heterogeneous units than nonwhites and women. This
finding raises the point that being different in a specific work setting is not the
whole story; there are also institutional or societal expectations about organiza-
tional composition and individual status, power, and rewards that affect how indi-
viduals react. Wharton has argued that "researchers should . . . examine how
organizational arrangements help to evoke . . . distinctions at both the individual
and group levels" (1992, p. 57) and her recourse to social identity theory linked
the study of what happens inside organizations to broader concerns of social cate-
gorization.
The importance of the social context for understanding the effects of de-
mography is nicely illustrated by a study of Japanese top management teams
(Wiersema and Bird, 1993). They found, first of all, that there was less diversity
in top management teams in Japan than there was, on average, in the United
States. This reflects Japanese organizations' policies of promotion from within,
slow evaluation and promotion, and an emphasis on consensual decision making.
But, even though there was less heterogeneity in Japanese top management
teams, the effect of what heterogeneity there was on turnover was substantially
stronger than that observed in studies of U.S. managers. Heterogeneity on the
dimensions of age, team tenure, and the prestige of the university attended were
all significantly correlated with team turnover. In Japan, there is less heterogeneity
in senior management, but heterogeneity also has a larger effect than in the
United States.
The distribution of length of service in an organization affects not only turn-
over and interpersonal processes such as conflict and social integration; it also
effects innovative activity and organizational change. At the group level of analy-
sis, R. Katz (1982) reported that research and development project groups with
either fairly short (less than one and a half years) or fairly long (greater than five
years) average tenure in the group were less productive than groups intermediate
in average tenure. He argued that this was because in groups with fairly short
86 New Directions for Organization Theory
tenures, individuals spent a lot of time organizing themselves and learning how
to work together and access necessary information and expertise distributed across
group members. In groups in which individuals had worked together on average
quite a long time, the negative effect on performance came from restricted com-
munication and hence less search and access to new information, particularly
from outside the particular project group, than might be optimal.
Much of demography's effect on organizational change and adaptation comes
through its impact on career processes in organizations. Stewman (1986) has
shown that "the commonly held view that organizational pyramids necessitate
declining career chances the higher one rises in the organization" is inaccurate
(Stewman, 1988, p. 175; see also Stewman and Konda, 1983). Rather, career
chances depend most importantly on the rate of organizational growth and to a
lesser extent on the rate of voluntary quitting. Thus, individuals in organizations
that cease growing face substantially more limited promotion prospects, and this
effect on careers has implications for attracting and retaining talent.
As a nice illustration of this phenomenon, S. Morris (1973) examined the
development and evolution of the U.S. railroad industry, one of the first industries
to have organizations operating on a national scale and one in which many of
the early management practices originated. However, some time after its creation
the industry stagnated, both in terms of its competitive success and in terms of its
management-systems innovation. Morris argued that much of the problem from
the 1920s onward was a consequence of the industry's demography. The railroads
grew quite rapidly around the turn of the century and then abruptly stopped
growing. This left them with a fairly young management cadre, but without the
business that created additional management positions, producing blocked pro-
motional opportunities for new organizational entrants. With few prospects for
mobility, new organizational entrants left, and recruiting of talent became more
difficult. When one considers that this was a period in which the average level of
education was rising, management practices were evolving as large companies
increasingly populated the landscape, and technological change was also acceler-
ating, the loss of expertise resulting from the inability to attract and retain young
managerial talent was substantial. Thus, a particular demographic situation gener-
ated personnel problems that left the organizations not well suited to solving the
performance problems, generating a cycle of decline that can be observed in
other industries as well.
Demography may make adapting to changing environmental conditions
more difficult. T. L. Reed (1978) studied the U.S. Foreign Service and its appar-
ent inability "to change rapidly in response to the shift in the complexity of inter-
national relations in World War II and thereafter" (Stewman, 1988, p. 188). Reed
argued that this problem arose from the organization's opportunity structure,
which was based on seniority and typically had long waiting times before some-
one could advance to the more senior grades. Because of this recruitment and
advancement process, there was "an extended time lag . . . to get new recruits
into the middle and senior levels of the expanded areas of responsibility" (Stew-
man, 1988, p. 188) and thereby bring new expertise and points of view into the
foreign policy development process.
Effects of Organizational Composition 87
Demography affects not only the rate of succession and adaptation but also
elements of the choice of a successor. In particular, Pfeffer and Moore (1980a)
found that academic departments that had more long-tenured individuals tended
to choose new department heads more frequently from within the department. If
for no other reason than availability, demography will affect succession — one is
more likely to choose a successor from within an organization if there is an abun-
dance of senior individuals with the requisite experience from which to choose.
Determinants of Demography
As Mitman has noted, most of the research literature on demography has explored
the effects of demographic distributions on various outcomes, and "only a small
number of studies have focused directly on the origins and determinants of demo-
graphic patterns" (1992, p. 5). Because organizational demography is determined
"by the past history of the social composition of net flows into it" (Stinchcombe,
McDill, and Walker, 1968, p. 221), understanding organizational demography
requires understanding demographic composition at some initial point and the
subsequent flows into and out of the organization. Because the processes of orga-
nizational entry, internal transitions such as promotion, and organizational exits
are themselves affected by demographic composition (e.g., the studies cited above
on turnover and the tendency for recruiting to reproduce friendship and social
composition), the evolution of an organization's demography over time is a dy-
namic process with important feedback loops. External labor market conditions,
government regulations such as antidiscrimination and affirmative action require-
ments and the vigor with which these are enforced, and economic factors such
as the rate of industry growth or decline are among those forces that affect entry
and exit processes and the consequent organizational demography. Although
modeling the dynamics over time imposes substantial data requirements, to the
extent demography is an important explanation for organizational processes and
outcomes, understanding demographic dynamics is an important undertaking.
One notable example of what might be done in the study of the determinants
of demography is a study of gender integration in the government of the State of
California (Baron, Mittman, and Newman, 1991). Using a linear partial-
adjustment model, the study examined the determinants both of the "target level"
of segregation — "a minimum level of gender segregation that is possible, given
the organizational configuration, toward which the agency is posed to adjust but
which social forces might resist" (p. 1376) —and the speed of adjustment to that
target level. The research employed quarterly staffing data over a six-year period
for all state agencies that described the ethnic and gender composition of each
job title within each agency—detailed data that permit quite fine-grained analysis.
Baron et al. found that both initial segregation levels and the speed of adjustment
were related to the same factors:
Segregation . . . was markedly lower in large agencies and those with high pro-
portions of female and nonwhite employees and was somewhat higher among
agencies that lacked an affirmative action program, were unionized along craft
88 New Directions for Organization Theory
lines, were clerical intensive, derived their mandate from the state constitution
|and consequently were presumably less subject to legislative influence], reported
objective quantitative output measures (i.e., had well-defined core technologies),
were not dependent on the General Fund for revenues, or experienced frequent
chief executive turnover, (p. 1385)
The study also found that agencies with female heads adjusted faster, as did agen-
cies that grew faster and faced more scrutiny. The effect of employment growth
on demographic change is likely to be an effect observed in many different orga-
nizational contexts, as is the importance of political interests, in this study repre-
sented by unions and the amount of legislative oversight and pressure from the
state personnel board. The methods and models employed in this study are
broadly applicable to the study of demographic change in general and should
have stimulated more follow-on studies than they did.
A second example of an analysis of changes in organizational demography
over time is Haveman's (1995) study of the entry rate, the exit rate, the average
tenure, and the dispersion in tenure for managers in savings and loans. She noted
that "explanations of organizations' tenure distributions must be rooted in explana-
tions of job mobility, because movement of employees into and out of the labor
force and movement of employees between organizations drives individual tenure
and thus drives changes in organizations' tenure distributions" (pp. 588—589).
Ilaveman found that the rate of entry into organizations was positively affected by
the industry merger rate, dissolution rate, and firm acquisition rate as well as by
the industry birth and dissolution rate. Average tenure was positively affected
by the industry dissolution rate and negatively by the firm acquisition rate. Tenure
dispersion was positively related to both the industry birth rate and the industry
merger rate. Haveman's study also indicated that larger and older organizations'
demographies are affected differentially by industry dynamics.
Most of the existing research on the determinants of organizational composi-
tion has focused on the proportion of women or minorities as the dependent
variable to be explained. Studying the gender composition of college and univer-
sity faculties, Tolbert and Oberfield (1991) found that there was a smaller propor-
tion of female faculty when there were more slack resources and that there was
more female faculty the higher the proportion of female students. This latter
finding is consistent with research by Borjas (1980; 1982; 1983), who explored the
connection between the proportion of minorities and women in constituencies for
agencies served by the federal government and the hiring and salary of women
and minorities in those agencies. He argued that to maximize political support,
agencies that served minorities and women would tend to match the demograph-
ics of their client population in their employees. He did find that the demo-
graphic composition of the groups served by agencies strongly affected both the
proportion of women and minorities hired and their relative wages (compared to
white men) —agencies serving women, for instance, employed a higher propor-
tion of women and the women's earnings were more similar to men's. Tolbert
and Oberfield's first result, on the effect of resource levels, echoes the results of
other studies that have also found that women tend to work in organizations that
Effects of Organizational Composition 89
have fewer resources (e.g., Szafran, 1984; Pfeffcr and Davis-Blake, 1987). Al-
though studying the determinants of gender and race composition of organiza-
tions is obviously important, it would be useful to have studies of other demo-
graphic factors such as age and tenure in the organization and, as in the Haveman
(1995) and Baron et al. studies described above, to have these changes dynami-
cally modeled.
At about the same time as Wagner et al. (1984) were studying how demographic
diversity affected turnover in top management teams, Mambrick and Mason
(1984) were proposing an "upper-echelons" theory of management that empha-
sized the organizational effects of the composition of the top management group.
Their argument was that both strategic choices and organizational performance
were related to the characteristics of the top management team because of the
presumption "that top managers structure decision situations to fit their view of
the world. . . . A central requirement for understanding organizational behavior
is to identify those factors that direct or orient executive attention" (Finkclstein
and Harnbrick, 1990, p. 484). Ilambrick and Mason argued:
a manager, or even an entire team of managers, cannot scan every aspect of the
organization and its environment. The manager's field of vision —those areas to
which attention is directed —is restricted, posing a sharp limitation on eventual
perceptions. Second, the manager's perceptions are further limited because one
selectively perceives only some of the phenomena included in the field of vision.
Finally, the bits of information selected for processing are interpreted through a
filter woven by one's cognitive base and values. (1984, p. 195)
This meant that the individual's worklvicw was probably related to his or
her functional background, education, and experience, all of which affected the
information one attended to and how one processed that information. Conse-
quently, research on the charactcristies of top management teams has emphasized
heterogeneity in functional backgrounds, the amount of time the team has spent
working together, and other indicators of experience, diversity and similarity. The
other important emphasis in this literature has been the focus on characteristics
of the team, as contrasted with the research linking strategy and organizational
performance to the characteristics of the chief executive (e.g., Carlson, 1972).
The evidence suggests that team effects are important in understanding organiza-
tional change and performance. Tushman, Newman, and Romanelli noted that
"the data are consistent across diverse industries and companies. . . . An execu-
tive team's ability to proactively initiate and implement frame-breaking change
and to manage convergent change seem to be important factors which discrimi-
nate between organizational renewal and greatness versus complacency and even-
tual decline" (1986, p. 15).
For instance, Finkelstein and Ilambrick (1990) examined 100 organizations
in the computer, chemical, and natural gas industries. They found that long-
90 New Directions for Organization Theory
tenured managerial teams tended to have more persistent strategies and that the
strategics and performance more closely conformed to industry norms. They also
found that the effects of top management team composition were strongest in
those competitive environments that permitted managers the greatest decision-
making discretion. This result is important in reminding us that any effects of top
management teams will be stronger in environments that permit managerial discre-
tion and in which performance variation is consequently greater. Bantel and Jackson
(1989) observed a negative relationship between top management team tenure and
organizational innovativeness. Wiersema and Bantel (1992), measuring corporate
strategic change as the absolute change in the level of diversification, found that
firms that had individuals in top management who were younger, had less average
organizational tenure, and had higher levels of education, more training in the sci-
ences, and longer tenure working together in the top management team tended to
experience more strategic ehange. Their results suggest that cognitive perspectives
are reflected in demographic measures of team composition.
O'Reilly, Snyder, and Boothe (1993) explored the effects of top management
demography in a sample of twenty-four electronics firms, defining the top man-
agement team as those who reported directly to the CEO. They measured turn-
over, organizational change, performance, and top management team dynamics
in terms of how positive these dynamics were using a seven-item scale. The study
controlled for company age, firm size, and the size of the top management team.
Consistent with other studies, they found that increased heterogeneity within the
team was negatively associated with team dynamics and was positively related to
turnover. They also found that "heterogeneity within the executive team is sig-
nificantly linked to less structural ehange, whereas homogeneity is positively asso-
ciated with change" (p. 167). They noted that "when an executive team operates
in a turbulent context, there may be an important premium on their ability to
function effectively as a group and to avoid conflicts and misunderstandings that
can slow them clown and deflect their attention" (p. 172). Thus, ironically, the
heterogeneity that brings different information and perspectives can also have dys-
functional effects, particularly if it is not managed effectively.
The O'Reilly et al. (1993) study is one of the few that have explored the
process through which top management team demography operates. In another
study of this process, K. G. Smith et al. (1994) used data from fifty-three high-
technology firms to test a direct model of the effect of top management team
demography on organizational performance, a model in which demography af-
fected performance through its influence on interpersonal processes, as well as a
model in which process itself had direct effects. Figure 4-1 shows the three ways
in which process and top management team demography might affect organiza-
tional performance. Controlling for past performance, firm size, industry demand,
and competition, K. C. Smith et al. (1994) found that demography affected in-
tervening processes such as informal communication, which in turn affected sales
growth and return on investment. They also observed direct demographic effects
of heterogeneous experience and heterogeneous education on these two measures
of performance as well as direct effects of informal communication and social
integration.
Figure 4-1 Models of the Effects of Demography on Performance. Source: K. G. Smith
ct al., 1994
These latter two studies are noteworthy because they are exceptions to much
of the upper-echelon literature that has been almost completely unconcerned
with processes of interaction among individuals. Recall that the organizational
demography literature focused on issues such as intcrcohort conflict, succession,
career advancement, and so forth —all of which are also issues within the top
management team. What most of the upper-echelon research has done, however,
is to simply aggregate individual characteristics — for example, average length of
tenure in the team or the organization, average education —but it has not ex-
plored the consequences of varying distributions of these characteristics for the
processes of conflict and career advancement that one would also expect to be
important in top management groups.
i)2 New Directions for Organization Theory
Virtually all of the top management team literature has focused on the conse-
quences of upper-echelon composition. That, of course, leaves the question of
how senior management composition emerges in the first place. Demographic
similarity may play an important role in understanding this process. Westphal and
Zajac (1995) examined 413 large corporations over the period 1986-1991. They
found that when the chief executive was more powerful, new directors appointed
to the board tended to be more similar to the existing CEO. When the board was
more powerful, new directors were demographically more similar to the existing
board members. They also found that the greater the demographic similarity be-
tween the chief executive and the board, the more highly compensated was the
CEO, even when other factors affecting compensation were statistically con-
trolled. Their study presents an upper-echelon case of the type of homosocial
reproduction —the tendency to hire and promote similar others —that Kanter
(1977) had described more generally.
Another study of the filling of senior-level college and university administra-
tive vacancies also illustrates the importance of demography for the construction
of senior management teams. The literature suggests that there is both a tendency
to choose similar others and a propensity for jobs to become categorized by gen-
der and race depending on their composition (Bielby and Baron, 1986). Konrad
and Pfeffer (1991) found that a woman or a member of a minority was more
likely to be placed in a vacant college administrative position (1) the higher the
proportion of women (or minorities) in similar positions in the labor market as a
whole; (2) the higher the proportion of women (or minorities) in senior adminis-
trative positions in the organi/ation; and (3) if a woman (or minority) had been
the previous occupant of the position. They found that each of these effects was
significant, even with the others in the model, which meant that the gender or
race of the position's previous occupant affected the gender or race of the new
occupant even with labor market supply and the power of the demographic group
in the senior administration statistically controlled. The study provided evidence
that positions were typed by the characteristics of the previous occupant and also
for the importance of the proportion of women in comparable positions in the
labor market and in the administrative hierarchy in the specific organization for
predicting the characteristics of those hired to fill vacant positions.
It is important to recognize that the top management teams literature, em-
phasizing the effects of senior team composition on organizational change, strat-
egy, and performance, represents a reaction to perspectives on both organizational
behavior and strategy that decrnphasize the role of action and choice:
John Child's (1972) article on strategic choice argued that political considerations
loomed large in understanding organizations and their structures. Child noted
that "many available contributions to a theory of organizational structure do not
incorporate the direct source of variation in formal structural arrangements,
namely the strategic decisions of those who have the power of structural initia-
tion" (p. 16).
The top management team literature provides evidence that backgrounds,
which, of course, represent individual history and learning, are important in un-
derstanding organizational dynamics. More important, it focuses attention on the
composition and interrelationships of the senior management team. This focus
stands, for instance, in contrast to the literature on leadership, which, for the
most part, examines the correlates and consequences of individual leader behav-
ior—even though we know that, outside of the experimental laboratory, leaders
seldom act alone in their roles. By demonstrating the importance of top manage-
ment team dynamics and the effects of composition on those dynamics, this
research has implications for constituting top management teams and for execu-
tive and leadership development, implications that remain to be more fully re-
searched and explored.
Ely (1995) argued that it was not just the proportion of women but where in
the hierarchy of power they were located that mattered. Using both quantitative
and qualitative data, she found that sex roles were more stereotyped and women
rated themselves less positively in firms with relatively low proportions of senior
women. But, Gutek, Cohen, and Konrad (1990), examining the sexual harass-
ment of women at work, found that the higher the proportion of women up to a
point, the more sexual behaviors and sexual harassment there was. The relation-
ship was curvilinear, because in an organization that was composed solely or
virtually entirely of women, there would be no sexual harassment. Their results
reflected the greater amount of contact between men and women when there
were more women in the work place.
An alternative perspective argues that women and minorities fare worse and
face more discrimination when they are in larger proportions (Blaloek, 1967).
The logic is that "a larger minority is thought to represent a greater perceived
threat to the economic and political security of the majority" (South et al., 1982,
p. 587). If there are only one or a few women or minorities, these individuals
themselves may face more pressure and less social support, but because they are
few in number, they may be different but are not a threat. In larger numbers,
they may be perceived as a threat or as competition, and this can engender more
forceful hostility on the part of the majority. "Members of the majority group are
assumed to respond to competitive threats posed by a large minority by actively
denying minority group members access to ... rewards" (p. 589). Writing about
the effects of increasing proportions of women, Allmendinger and Hackman
noted, "The Kanter position is that things will get better, mainly because of what
the women are doing; the Blaloek position is that things will get worse, mainly
because of how the men are reacting" (1995, p. 427).
South and his colleagues administered a survey to employees in six offices of
a government agency. They examined the effect of the gender composition of the
work group on the frequency of contact with co-workers and the amount
of encouragement for promotion from male and female co-workers as well as
from a male supervisor. The results showed that there was less contact with male
co-workers and more contact with female co-workers the higher the proportion of
women in the work group, but there was no effect on the amount of contact with
a male supervisor. There was no statistically significant effect of group gender
composition on promotion encouragement from male co-workers, but the higher
the proportion of women in the work group, the less encouragement there was
from both the male supervisor and from female co-workers. South et al. (1982)
argued that both contact and competition effects were important, and which was
Effects of Organizational Composition 95
stronger depended not only on the particular proportions but also on the specific
context. In another study of this issue, Toren (1990) found that the greater the
proportion of women on the faculty of Israeli universities, the lower were their
academic ranks relative to those of their male colleagues. And, Wharton and
Baron (1987) reported that men in mixed-gender work settings had lower levels
of job satisfaction.
Allmendinger and Hackman (1995) studied seventy-eight symphony orches-
tras in four countries that varied in the proportion of women members. This is a
particularly appropriate setting, because "variation associated with both organiza-
tional tasks and member talent is relatively low across both orchestras and na-
tions" (p. 428), eliminating alternative interpretations for the results of gender
composition. The study collected questionnaire data from a sample of orchestra
members on a number of dimensions including (1) the organizational features of
the orchestra, such as integrity as an ensemble, player involvement, and recogni-
tion; (2) behavior of the music director; (3) orchestral processes, such as the qual-
ity of relationships and musical outcomes; (4) player motivation and satisfaction;
and (5) specific satisfaction with features such as management, work relationships,
and job security. They found that gender composition was significantly related to
nineteen of twenty-two dependent variables, with there being less effect on the
music director's behavior. "Almost all of the dependent measures decline signifi-
cantly as the proportion of women in an orchestra increases from token to more
substantial levels. However, 10 of the 19 significant effects exhibit a statistically
significant reversal of the decline at some point on the gender composition spec-
trum" (p. 435). Moreover, ratings by observers of two features, integrity as an en-
semble and orchestra structure, matched the members' survey descriptions and also
evidenced a negative effect for increasing proportions of women (p. 443). The
study also found that it was "men who respond most strongly to changes in organi-
zational gender composition" (p. 443). The stirdy is consistent with the view that
"things really do get more difficult . . . as more women enter traditionally male
organizations, at least until a state of gender balance is achieved" (p. 454). The
importance of this line of work is to note that there are competitive effects from
increasing the proportions of those in token status in a work context, and these
need to be considered along with the effects of contact and social support.
Konrad, Winter, and Gutck have drawn a useful distinction between theories
that consider proportions and numbers by themselves and those that bring in the
status characteristics of those in greater or lesser proportion as being important in
understanding organizational dynamics:
"Generic" composition theories assume that the important factor affecting indi-
viduals in work groups is whether they are in the majority or in the minority and
that the specific demographic characteristics of the majority and the minority do
not add anything to the prediction of individual behavior and outcomes. . . .
"Institutional" composition theories assume that members of different demo-
graphic groups will react differently to being in the minority or the majority,
depending upon the social significance of particular demographic characteristics
in society at large and in the particular organization. (1992, pp. 116-117)
96 New Directions for Organization Theory
Their data from eighty-nine work groups provided support for an "institutional"
version of the effects of composition. The authors found that "women holding
superior positions in predominantly male work groups experienced the most un-
desirable social outcomes of any group" (p. 116).
Although women themselves sometimes may do better in larger proportions,
organizations and occupations in which they constitute a larger proportion fre-
quently offer fewer rewards. For example, the proportion of women administrators
in colleges and universities has been found to be negatively related to the salaries
earned by both men and women administrators (Pfcffer and Davis-Blake, 1987).
Baron and Newman (1990), studying the California Civil Service, reported that
the proportion of women and minorities in a given job was negatively related to
the prescribed salary for that job, even when education and experience require-
ments were controlled, and this effect helped explain changes in prescribed wage
rates for positions in the state civil service over time. Shenhav and Haberfcld
(1992) found that both men and women scientists in disciplines employing a
higher proportion of women had lower earnings. England et al. (1988) have re-
viewed the extensive evidence demonstrating that, at the occupational level of
analysis, wages are negatively related to the proportion of women in the occupa-
tion.
Pfeffcr and Davis-Blake's study of academic administrators observed an im-
portant nonlinearity in the relationship between the proportion of women and
wages. At very low and at very high percentages of women, when the addition of
a few more women would not make a great deal of difference in how the organi-
zation and its work was seen, the negative effect of an increase in the proportion
of women on salaries was less than around a tipping point, a value at which
work apparently became reelassified as being "women's work" and consequently
devalued.
The organization- or job-level effects of the proportion of women on wages
probably occur because of the societal devaluation of women's work. It is im-
portant to note that these results eontradiet human capital theory, with its empha-
sis on individual abilities and experience, for they demonstrate a compositional
effect on economic outcomes net of individual differences.
What may be of equal interest, however, are the more micro-level processes
occurring within organizations that affect women's progress. These processes in-
volve the availability of similar others with whom to form close relationships and
the trade-off between network relations that provide social support through
stronger ties and network relations that provide access to a wider range of informa-
tion and contacts but entail weaker tics.
Because there are, in most organizations, fewer women (and minorities) in
high-status positions in the hierarchy, in most contexts men and women differ in
the opportunity to have informal interaction with same-gender others who arc
also instrumental!}' useful (Ibarra, 1992; 1993b). Because there are preferences
for same-gender relationships, particularly for providing the social support that
requires strong social ties, women have social networks and task-related, instru-
mental networks that overlap less than men's. Thus, the difference in men's and
women's career attainment is not simply that men and women have different
Effects of Organizational Composition 97
human capital variables, such as tenure and graduate education, and positional
resources, such as rank and potential for advancement, were held constant, mi-
nority and white middle managers in comparable jobs differed in the homophily
and intimacy of their organizational network ties" (1995, p. 693), with minority
ties being less homophilous and close. This evidence is consistent with structural
constraints on network patterns, although the study also found strategic differ-
ences in network behavior that differentiated high-potential from other minority
managers.
There is compelling evidence that network relations predict career success
(e.g., Hurt, 1992). This has led to discussions of the concept of social capital, to
complement ideas such as human capital and cultural capital. Burt defined social
capital as "at once the resources contacts hold and the structure of contacts in a
network" (1992, p. 12). He distinguished social capital from human capital by
noting that human capital is a property of an individual while social capital is
"owned" jointly by the parties to a relationship. If the relationship is severed, one's
social capital that originates in that relationship disappears. An individual's social
capital depends on his or her structural position in the social system. Although
there may be some ability to strategically affeet one's social relations, this ability
is clearly constrained by, among other things, the social composition of the envi-
ronment in which one is operating. Also, building network ties and maintaining
social relations takes time, and thus investment in networking must be balanced
against alternative activities. The fact that studies of career attainment are able to
explain variance from a knowledge of network relations provides support for the
social model of behavior and for the importance of the social context in under-
standing organizations.
Conclusion
behavior, with its emphasis on social influence, which often operates through
interpersonal networks or structural equivalence. Composition helps to determine
the content of those influences and the composition of the networks through
which social influence travels.
In a fashion similar to other areas of organization studies, there has been a
tendency to overlook the structural context within which individual behavior oc-
curs. The leadership literature is a good example of this, in its neglect of the
effect of managerial team composition on leader behavior. What the research on
demography and social context has done is to bring these issues of structural
constraint back into focus. A large number of substantively important empirical
contributions suggest that one ignores the effects of social composition at some
peril. Demography has a tendency to perpetuate itself—to use Kanter's (1977) apt
phrase, to exhibit "homosocial reproduction." And demography has effects on
organizations ranging from innovation and change to strategic decision making to
the welfare of women and minorities in those organizations. As such, the study of
organizational demography reminds us yet again of the social, relational nature of
organizations and has provided both theoretical ideas and some empirical appara-
tus for taking these ideas seriously.
5
The various models of behavior have quite different implications for how to
best achieve the requisite level of social control and coordination, and assump-
tions about human behavior inevitably undcrgird all organizational control sys-
100
Mechanisms of Social Control 101
terns. What is often not done, either by the theorists or by organizational manag-
ers, is to make these assumptions explicit. O'Reilly and Chatman provided a good
definition of control in organizational settings: "control comes from the knowl-
edge that someone who matters to us is paying close attention to what we are
doing and will tell us if our behavior is appropriate or inappropriate" (1996,
p. 161).
The control process begins with recruiting and selection, in which organiza-
tions try to find individuals who already possess necessary skills and attitudes that
make them good prospects for being productive. Uncovering who is a good em-
ployment prospect and possesses the requisite qualities is inevitably an uncertain
undertaking. One cannot always directly observe motivation, diligence, or intelli-
gence. Consequently, in the recruiting process, organizations frequently rely on
signals or proxies for the underlying quality being sought.
ing (Pascale, 1985) to help them understand what is expected and to come to
want to do those activities. Individuals also receive wages, the possibility of promo-
tion, and other benefits — incentives designed to induce them to remain in the
organization and to work hard in its interests. Also, people work most often in
hierarchies under supervision. How well they are led may also affect their atti-
tudes and behavior.
Because control and coordination are such ubiquitous issues, virtually the
entire organizations literature concerns this topic, at least indirectly. This chapter
discusses four mechanisms of social control: (1) the use of rewards and incentives,
including "negative" rewards and surveillance; (2) commitment and socialization
processes; (3) organizational culture; and (4) leadership. Each of these processes
is a fundamental mechanism of social control in organizational settings and is
intended to direct and motivate behavior, and each has stimulated substantial
research and theoretical interest and controversy.
There is an important difference in emphasis across these mechanisms of
social control. Rewards and incentives work primarily on the environment of em-
ployees, and the literature on organizational rewards almost invariably proceeds
from the presumptions of the economic model of behavior. As such, reviewing
the literature on rewards and compensation will permit us to see in yet another
research context to what extent the predictions of the economic model of behav-
ior are, in fact, empirically supported and under what circumstances. The use of
rewards and incentives as mechanisms of organizational control varies over time
and place depending on the degree of acceptance and dominance of the eco-
nomic perspective on organizational behavior. By contrast, commitment, culture,
and some aspects of leadership try to engage and change fundamental decision
premises, self-perceptions, beliefs, and values and frequently work on the emo-
tions using social psychological principles, using the heart as well as the head. As
such, these control strategies are premised largely on the retrospectively rational
and, to a lesser extent, the social and moral models of behavior.
Interest in rewards and incentives —a rational basis of control—and norma-
tive control, based importantly on social psychological principles, has waxed and
waned over time and, contrary to some accounts (Edwards, 1979), has not simply
evolved from coercive to normative control (Barley and Kunda, 1992). Table 5-1
reproduces Barley and Kunda's depiction of the succession of different managerial
ideologies over time. These bases of control have fluctuated over time because of
the fundamental tension between community and individual self-interest that
forms the basis for the various control systems:
each sensed that industrialization was problematic because it juxtaposed two con-
trasting paradigms for social order. These two forms of social organization were
given different names by different scholars. Weber wrote of the "communal" and
the "associative." Durkheim contrasted "mechanistic" with "organic" solidarity.
. . . However, the essence of their vision was the same. In a Gemeinschaft, people
share a common identity, are bound by common values and traditions, and par-
take of a way of life that contrasts sharply with the competition, individualism,
and calculative self-interest associated with the Gesellschaft. The central di-
lemma identified . . . concerned the integrity of the social fabric. How could
relations based on utilitarianism and rational calculation remain integrated and
socially fulfilling? (p. 385)
There are a number of fundamental issues in the literature on rewards and incen-
tives:
1. What determines the incentive structure organizations actually em-
ploy with different groups of people? For instance, why is there pay
for performance on an individual basis in some organizations but pay
for seniority in others? Within a given organization, why are some
jobs paid more on the basis of individual performance than others?
2. What are the effects of various amounts, forms, and distributions
104 IVew Directions for Organization Theory
Throughout the discussion that follows, I focus on pay, which is the principal
(although not the only) reward in organizations. But A. Kohn (1993) argues that
the issue isn't what is given as a reward but rather the fact that something-
praise, a prize, or pay—is provided contingent on the focal person exhibiting
some desired behavior. Thus, the issues and principles discussed are general, but
the focus on pay has the advantage of more directly engaging a lot of the empiri-
cal literature and confronting the predictions of the economic model most di-
rectly. Although there has recently been much more empirical research on pay
practices in organizations, stimulated by the availability of data sets on managerial
compensation collected by various compensation consulting firms, definitive an-
swers to any of the questions posed must await further research.
There are several facts one needs to keep in mind in thinking about these
issues. First, the bases of rewards are changing, particularly in the United States
but almost certainly elsewhere as well. One basic dimension of the change is the
growing emphasis on variable pay—pay that changes depending on the perfor-
mance (or skill) of the individual, his or her work group, or organization and a
corresponding dcemphasis on paying simply for time worked or for seniority. The
change can be seen most dramatically in the pay of senior managers such as chief
executive officers. Useem (1996), studying the compensation of the seven top
executives from forty-five large companies over the period 1982-1995, reported
that the proportion of managers' total compensation that was variable increased
from 37 percent to 61 percent. The proportion of variable pay for CEOs in-
creased even more, from 41 percent in 1982 to 70 percent, in 1995. The biggest
change was an increase in the use of long-term incentive pay.
Nor is the increase in variable pay plans confined to senior management.
Ledford, Lawler, and Mohrman (1995) reported the results from surveys of the
pay practices of the 1'brtune 1000 (the 500 largest industrial and 500 largest ser-
vice firms) conducted in 1987, 1990, and 1993. Between 1987 and 1993, the
proportion of firms using individual incentives for at least 20 percent of their work
force increased from 38 percent to 50 percent; the proportion using gain sharing
for at least some of the work force increased from 7 percent to 16 percent; and
the proportion of firms that had an employee stock ownership plan covering at
least 20 percent of the work force increased from 53 percent to 62 percent. The
proportion of firms using profit sharing for some of the work force actually de-
clined slightly over the period, from 45 percent to 43 percent. Although the
growth in some of these innovations in incentives is modest, there is little doubt
that there is much more experimentation occurring with incentives throughout
Mechanisms of Social Control 105
organizations (Kanter, 1987), with the emphasis being to link pay to performance,
thereby making pay more contingent or variable. Increased use of variable pay is
also evident in other countries besides the United States. Wood, in a study of 135
manufacturing plants in the United Kingdom, reported that "some form of merit
pay was claimed to be used in 50 per cent of plants, and profit-sharing in 25 per
cent, and moreover, the adoption of both increased between 1986 and 1990, by
13 per cent in the case of merit pay and 8 per cent in the case of profit-sharing"
(1996, p. 60).
This movement is quite consistent with operant conditioning principles (e.g.,
Skinner, 1953; Luthans and Kreitner, 1975), which emphasize the connection
between actions and their consequences. These changes are congruent with the
economic model of behavior, with its emphasis on incentives, particularly exter-
nally controlled and administered incentives, as being important. Indeed, one of
the important challenges to the economic model has been the historically rela-
tively small relationship between pay and performance, even for senior-level exec-
utives (Baker, Jensen, and Murphy, 1988). Jensen and Murphy noted, "actual
compensation contracts look very different form those predicted by economic the-
ory. The failure of theory to explain actual compensation arrangements suggests
that the theory is wrong" (1987, p. 43).
Second, reward practices vary dramatically across organizations and even
more dramatically across cultures. There are large, well-known, and effective
firms (e.g., Mars Candy, Southwest Airlines) that use virtually no merit pay. There
are others, such as Hewlett-Packard, that do use individual merit pay. There are
countries, with Japan and Korea being notable cases, in which pay is based pri-
marily on seniority, to a lesser extent on position, and virtually not at all on
individual differences in merit or performance, although these practices may be
changing. There are companies and countries in which pay is emphasized as the
primary mechanism for controlling and directing behavior, and there are coun-
tries and companies in which pay is deemphasizcd, with control being achieved
in other ways and rewards and recognition being more symbolic than economic.
The distribution of rewards also varies dramatically across companies and cul-
tures, with large dispersion in rewards characteristic of some but not others.
This substantial variation in pay practices is interesting and important for two
reasons. First, if some pay practices were invariably more effective than others
regardless of the context, one would expect to see more uniform adoption of these
more effective practices as a response to competitive pressure. Second, because
of the strong ideological component of pay—practices are frequently based more
on beliefs about behavior than on hard evidence (A. Kohn, 1993)—this is a topic
in which there is much more good theory than field-based empirical results that
inform the theory. Thus, there is an important opportunity for research progress.
alternative pay systems in the American labor market is a result of a complex set
of economic, historical, and institutional forces. A folk wisdom has grown up
around the different types of plans" (1990, p. 87). Much of what we do know
about pay and incentive schemes is, however, at variance with predictions from
the economic model of behavior.
We know that pay is less differentiated in union as contrasted with nonunion
settings and that pay is more strongly related to seniority and is seldom related to
individual performance differences in unionized work places (Freeman and Med-
off, 1984). We know that even in nonunion private corporations that have perfor-
mance appraisals and presumably some form of merit pay, pay is often signifi-
cantly related to seniority and is not much related to performance if at all (Mecloff
and Abraham, 1980; 1981). There are substantial gender and race effects on pay
(Jacobs, 1992; Johnson, 1978; Sigelman, Milward, and Shepard, 1982), evidence
of discrimination in the wage determination process and that economic returns
are affected by more than an individual's human capital.
Leonard (1990), in a study of about 20,000 executives, found that pay was
related to the degree of organizational hierarchy—executives in flatter hierarchies
earned less. Lambert, Larcker, and Weigelt (1993), studying more than 4,000
executives in 303 different companies, found that the relationship between pay
and other factors such as performance, the number of people supervised, or chief
executive power varied greatly across industries. This suggests that managerial
labor markets are not completely efficient and that there are important industry
(and undoubtedly company and national) differences in the wage determination
process.
In a challenge to strictly economic, efficiency-based conceptions of the pay
determination process, Lambert and his colleagues found effects of managerial
power on pay. For instance, they reported a positive association between the level
of executive compensation and the percentage of the external board members
who had been appointed by the chief executive; they also reported statistically
significant effects for virtually all of their other measures of managerial power
such as the percentage of stock owned by the chief executive. Other evidence
also suggests the importance of power in the salary determination process. A study-
by Pfeffer and Konrad (1991) of academic salaries found that power had a main
effect on pay and positively affected the returns to research productivity and years
of experience, with more powerful individuals enjoying greater returns. Several
indicators of individual power were employed in that study: (1) receiving outside
research funding, (2) having an outside job offer, (3) extensive contact and com-
munication with individuals in other institutions, and (4) a scale composed of
responses to questions asking about the level of involvement in departmental af-
fairs and perceptions of opportunities to influence policies in the department and
in the broader institution. Moore and Pfeffer (1980) found that more powerful
departments on two campuses of the University of California were able to achieve
greater acceleration in salaries for their faculty members even when measures of
departmental quality and size were controlled.
A study of Australian establishments (Drago and Heyvvood, 1995) explored
Mechanisms of Social Control 107
predictions from the so-ealled new economics of personnel literature (Hart and
Holmstrom, 1987), which relates pay schemes to the ease of monitoring worker
effort, as well as predictions from the managerial strategic choice literature (Ko-
chan, Katz, and McKersie, 1986). The reported results for the use of various pay
systems such as piece rate, group pay, and profit sharing were both complex and
did not consistently support either theoretical perspective. For instance, the use
of piece rates was positively related to the proportion of long-tenured employees.
This result is exactly opposite to the predictions of the new economics of person-
nel literature, which argues that "firms will tend to avoid more immediate wage
incentives, such as piece rates, where the expected job tenure of workers is long,
since increasing wages over the employees' life-cycle provides a less costly alterna-
tive" (Drago and Heywood, 1995, p. 508). They also reported that the use of
piece rates was higher the larger the proportion of managers, which contradicts
the prediction that individual incentives provide an alternative to direct supervi-
sion. And they reported that the use of piece rates was less the larger the propor-
tion of casual labor—but one would expect that it would be casual workers, with
no long-term attachment to the organization or their fellow workers, who would
most require individual incentives to induce effort. They did find, consistent with
the strategic choice-managerial strategy predictions, that there was more use of
profit sharing in organizations that had quality circles or other systems to encour-
age worker participation.
Wood's (1996) study of U.K. manufacturing plants found no statistically sig-
nificant relationship between the use of high-commitment work practices and any
particular payment system. His study did find that the proportion of pay com-
prised of profit sharing payments was associated with the use of high-commitment
work practices, which means that "when profit-sharing is taken seriously, it is
likely to be in conjunction with" high-commitment management (p. 64).
We know that pay is strongly related to performance in professional sports
(e.g., Harder, 1992), but even here there is some evidence of racial discrimination
(Kahn and Sherer, 1988; Scully, 1974). Of course, performance is quite measur-
able in professional sports, and, moreover, there is an efficient national labor
market (for experienced players who can change teams) that makes the strong
pay—performance connection expected.
From a study of college and university faculty (Konrad and Pfeffer, 1990), we
know that research productivity, a measure of performance, had a larger effect on
pay in private, higher quality institutions, in institutions governed by collective
bargaining agreements, in institutions in which there was more collaboration on
research and more social contact among faculty, in departments that had chair-
persons with shorter, fixed-length terms and had norms emphasizing research,
and in fields with more highly developed scientific paradigms. However, even in
contexts in which productivity could be readily assessed and in which merit was
ostensibly emphasized, the effect of productivity on salary was quite small. Being
one standard deviation above the mean for research productivity produced only a
5.5 percent increment in salary. The results suggest that paying for productivity is
more likely where productivity can be more unambiguously assessed (fields with
108 New Directions for Organization Theory
Consider those individuals who were in settings that were above the median,
simultaneously, on resource levels, paradigm development, research emphasis,
research collaboration, and who worked in high quality institutions. . . . Individ-
uals producing research at a rate that is one standard deviation above the mean
in their fields lost more in salaries if they were in the wrong organizational con-
text than they would have lost had they simply worked less hard and produced
less research, falling to the mean level. (1990, p. 277)
argued that when teamwork and cooperation are important, individually based
rewards can reduce the incentive to cooperate. Milgrom and Roberts (1988) have
argued that merit systems that provide very different rewards can encourage un-
productive political behaviors as individuals jockey for those rewards.
Studying this issue empirically, Pfeffer and Davis-Blake (1992) found that
salary differentiation was related to turnover among academic administrators, with
those lower in the salary distribution being particularly affected by pay dispersion.
Pfeffer and Langton (1993), studying college and university faculty, found that
the greater the pay dispersion, the lower the job satisfaction, arid that paying for
performance and greater pay dispersion were associated with diminished levels of
research productivity and with a smaller likelihood of working collaboratively on
research. They noted that such adverse effects of wage dispersion were particularly
pronounced for individuals with shorter tenure (who were less committed), in
public colleges (where pay is more likely to be known), and in departments that
had less highly developed scientific paradigms (there was less consensus on stan-
dards of evaluation and performance). Cowherd and Levine (1992), using a sam-
ple of 102 business units, compared the pay and inputs of lower-level employees
to those of top management to compute a measure of pay equity. They found
that the greater the degree of pay equity across classes of employees, the higher
was the level of product quality. Thus, one answer as to why there is not more
use of differentiated incentives is that at least in some contexts they have negative
consequences.
Many of the studies of pay determination process have focused on CEO pay.
The basic message of these studies is that, at least historically, there has been
either no or a very weak relationship between CEO pay and company perfor-
mance, where performance is measured in almost any way (Crystal, 1991; Main,
O'Reilly, and Wade, 1995; Miller, 1995). Instead, CEO pay seems to be related
to organizational size, to the industry in which one works—with there being a
premium for working in unregulated and more competitive and risky industries —
and to the composition of the compensation committee of the board of directors.
O'Reilly, Main, and Crystal (1988) found that there were important social influ-
ences in the pay determination process (see also Main, O'Reilly, and Wade,
1995). Specifically, CEOs in companies in which members of the compensation
committee were themselves more highly compensated in their own roles as CEOs
earned more, controlling for other factors —a result the authors interpreted as
meaning that salaries were set in a process of social comparison. Compensation
committee members who themselves were well paid would be more generous
when setting the salaries of others because those salaries would not look high by
comparison. O'Reilly and his colleagues also discovered that this result was partic-
ularly strong when the compensation committee directors had joined the board
after the CEO had assumed his position —which probably meant that they owed
their positions to the CEO and therefore were more under his influence. Main
ct al. (1995) further reported that demographic similarity between CEOs and
members of the compensation committee, an important basis of interpersonal
attraction (Berscheid and Walster, 1969), was positively related to CEO pay.
The literature on CEO pay has attempted to provide an explanation for the
no New Directions for Organization Theory
increasing difference in pay between CEOs and the rest of the organization —an
explanation that is consistent with principles of economic efficiency—by devel-
oping a tournament model of executive compensation. The theoretical problem
is nicely described by La/ear and Rosen (1981, p. 847): "On the day that a given
individual is promoted from vice-president to president, his salary may triple. It is
difficult to argue diat his skills have tripled in that one-day period, presenting
difficulties for standard theory" that relates wages to human capital or to produc-
tivity. The basic idea behind the tournament model is that vice presidents (or
those one level below the CEO) compete for the top job. In order to induce
maximum effort, the "prize"—the salary earned when one achieves that posi-
tion—must be sufficient to compensate for the effort and the probability of
achieving the prize. So, Rosen (1986) "has shown that top prizes of a dispropor-
tionate size are theoretically necessary to motivate tournament survivors so they
will not rest on past achievements or winning as they enter the final contest.
Elevating the top prize has the effect of lengthening the career ladder for high-
ranking contestants" (O'Reilly et al., 1988, p. 260).
The predictions of the tournament model arc straightforward—the lower the
probability of winning the contest, or the greater the number of contestants, the
greater the difference in salary should be to compensate for this reduced expected
payoff. This model would suggest, for instance, that as organizations have flat-
tened their hierarchies, salary differences across levels would be expected to in-
crease to induce effort on the part of those who now have a lower expectation of
being advanced because of the greater number of people competing for the fewer
number of promotions. In one empirical test of this model, its predictions were
found not to hold. O'Reilly et al. (1988, p. 266) reported that, controlling for
other factors, the number of vice presidents was negatively related to CEO com-
pensation (and to the difference between the salaries of CEOs and vice presi-
dents). Moreover, also inconsistent with tournament ideas, the average vice-
presidential salary was positively related to CEO salary. The fact that the tourna-
ment model is inconsistent with the evidence has, unfortunately, not diminished
its popularity as an explanation for the rise in CEO compensation.
It is possible and indeed likely that because of the prevalence, particularly in
the United States, of belief in the economic model and its associated assumptions,
pay systems may be designed for legitimation —to signal to outsiders that the firm
is well managed and doing the right thing by implementing incentives to tie
senior executive pay to performance. Westphal and Zajac (1994) studied 570 of
the largest corporations over a two-decade period, examining not only which firms
adopted long-term incentive plans but which firms actually used them. They dis-
covered that a substantial fraction of firms announcing the adoption of long-term
incentive schemes in their proxy statements never actually used such plans to set
compensation and that this separation between form and substance was particu-
larly likely in organizations that had poor prior performance and powerful chief
executives. This result suggests that pay and pay systems have important symbolic
and legitimating aspects.
Mechanisms of Social Control 111
tingent pay or pay for performance can be based on individual performance, sub-
unit or work-group performance, the performance of the entire organization, or
some combination. Wclbourne and Cable have noted that "very little is known
about how employees covered by various group incentives interpret the programs"
(1995, p. 712). They also noted the absence of literature on the effects of
group incentives on job satisfaction. There have been studies of some forms of
group incentives, such as gainsharing, stock ownership, and profit sharing on
organizational-level outcomes (e.g., Kaufman, 1992; Weitzman and Kruse, 1990;
Rosen, Klein, and Young, 1986; Jones and Kato, 1993). There has not been much
study of the effects of group or organizational incentives on individual behavior
and attitudes. The distinction among the unit of analysis rewarded should be kept
in mind in thinking about incentives and behavior.
It is also the case that wage incentives are effective only to the extent that
the measurement system that underlies them measures the right thing. Most in-
centive systems are premised on the idea of individual merit and confront the
problem of interdependence. They all too often solve this problem by measuring
the wrong thing, and not very well. Simon provided a compelling critique of the
typical pay scheme and why it would probably not be effective:
If the indices used to measure outcomes are inappropriate, either because they
do not measure the right variables, or because they do not properly identify
individual contributions, then reward systems can be grossly inefficient or even
counterproductive. . . . In general, the greater the interdependence among vari-
ous members of the organization, the more difficult it is to measure their separate
contributions. . . . But of course, intense interdependence is precisely what
makes it advantageous to organize people instead of depending wholly on market
transactions. . . . Although economic rewards play an important part in securing
adherence to organizational goals and management authority, they are limited in
their effectiveness. Organizations would he far less effective systems than they
actually arc if such rewards were the only means, or even the principal means,
of motivation available. (1991, pp. 33-34)
As implied by the foregoing discussion, the evidence on the effects of varia-
tions in pay policies in organizations on performance is mixed. Leonard (1990)
found that firms with long-term incentive plans enjoyed greater increases in re-
turn on equity over the period 1981-1985, but he also reported that corporate
success was not related to the level of executive pay or to the steepness of the pay
differentials across ranks —a measure of the reward for achieving promotion.
Abowd (1990), in a study of 16,000 managers, found that the sensitivity of mana-
gerial compensation to performance was related to measures of shareholder return
but not to accounting measures of performance. This result is somewhat surpris-
ing, since executive pay tends to be set more frequently on the basis of the
achievement of objectives measured by the accounting system. Asch (1990), ex-
amining an incentive program for navy recruiters, found a positive effect of finan-
cial incentives, but there were also some negative results. The best recruiters
(based on historical data) produced less as they approached the prize eligibility
date, and after winning a prize, recruiters enlisted fewer recruits. Studying manag-
ers in a single company, Kahn and Sherer (1990) reported that managers for
Mechanisms of Social Control 113
whom bonuses were most sensitive to performance performed better the following
year, even when past performance was statistically controlled. However, the same
study found that differences in the sensitivity of merit pay to performance had no
significant effects on subsequent performance —thus, bonuses had an effect on
performance but merit pay did not.
A longitudinal study tracing the implementation of merit pay in a govern-
ment bureaucracy reported no relationship between merit pay and performance
(Pearce, Stevenson, and Perry, 1985). Marsden and Richardson (1994), studying
the use of individual performance pay in the Inland Revenue department in the
United Kingdom, reported that performance pay had little motivational effect and
may have demotivated some employees.
In a study of 8,000 workers in the United States and Japan, Levine (1993)
found that employees receiving high wages were less likely to quit, were more
satisfied with their pay, and reported a higher level of effort on the job. These
results support the idea that paying above-market wages may be rational because
it results in lower turnover costs and more employee effort, but Levine's data do
not speak to the issue of the effects of contingent pay practices.
The principal controversy in the pay literature concerns the use of individual
incentives (as contrasted with group or organizational incentives), and particularly
the effectiveness of contingent or merit pay. Deming and many other of the qual-
ity gurus have bemoaned the use of individual merit pay, arguing that it fre-
quently rewards or punishes workers for things over which they have little control,
since they don't often control the system of production within which they work
(Gabor, 1990). They also maintained that merit pay practiees often leave workers
discouraged and dissatisfied and create internal organizational competition and
conflict.
Other authors have noted that individual merit pay practices are inconsistent
with many of the underlying precepts of high-commitment work practices. For
instance, Beer ct al. maintained that "by making pay contingent upon perfor-
mance (as judged by management), management is signaling that it is they—not
the individual—who arc in control" (1984, p. 114). Consequently, "performance-
related pay may lower the individual's feeling of competence and self-determina-
tion and run counter to an intrinsic reward policy" (Wood, 1996, p. 55). Because
"out-put based incentive schemes, and particularly piecework, arc . . . very much
identified . . . with the past emphasis on a Taylorist control strategy and the ills
associated with it" (p. 56), many argue that piecework and other individual incen-
tive schemes conflict with the ideas of commitment, involvement, and intrinsic
motivation.
Besser's study of pay practices at Toyota Motor Manufacturing in Kentucky
found evidence of high productivity, quality, and discretionary effort in a work
setting in which equal pay was the norm — "81 percent of T'MM's employees
receive roughly the same pay with distinctions based on job category, not individ-
ual performance" (1995, p. 388) Bcsser noted that most motivation theories (e.g.,
Lawler, 1981) focused on the individual and took what motivated the particular
individual as given. By contrast, "Cole's (1979) contention and the argument of
this paper is that large Japanese firms take a much more active role in shaping
114 ^ew Directions for Organization Theory
employee motivation than organizations are usually given credit for. They do this
. . . by turning small work groups and company ideology into official reward
mechanisms with powerful consequences for individual motivation" (p. 387). Acl-
ler's (1992) study of New United Motor Manufacturing also documented an in-
stance of extremely limited use of differential monetary rewards in a system that
was highly effective by any measure.
A good portion of the literature on the effects of incentives has focused on
the connection between CEO incentive schemes and organizational perfor-
mance. This research is frequently plagued by two problems. One is that the
choice of incentives is not exogenous. For instance, a CEO who anticipates im-
proved corporate performance and a rising stock price is much more likely to
want to be paid with stock options or incentive compensation. Thus, a study that
simply finds a contemporaneous correlation between incentive compensation and
performance may have the causality reversed—the incentives may be imple-
mented because of the enhanced performance (Leonard, 1990). Second, the as-
sumptions that CEOs, already exceedingly well compensated, will work harder
for some extra compensation and that a single individual, even a CEO, can pro-
foundly affect the performance of large organizations are both suspect. There is
little evidence, for instance, that CEOs who have long-term incentives as part of
their compensation package in fact manage companies who do better in terms of
economic returns over time (Crystal, 1991; but, see Leonard, 1990).
There are more consistent results on the effects of incentives in studies of
sports. There is, for instance, evidence that larger prizes produce better scores in
golf tournaments (Ehrenberg and Bognanno, 1988; 1990), even controlling for
the quality of the players who choose to participate. A study of race car drivers
found that larger prizes produced greater speed but also more accidents (Becker
and Hnselid, 1992). Of course, many sports, including golf and race car driving,
involve comparatively little interdependence among the contestants. This absence
of important task-related interdependence is the exception, rather than the rule,
in most work organizations. Consequently, results from studies of incentive effects
in sports need to be generalized to other settings with some caution.
stress (Amick and Smith, 1992; Smith, et al., 1992; Irving, Higgins, and Safayeni,
1986), decreased job satisfaction (Irving, et al., 1986; Grant and Higgins, 1989),
increased feelings of social isolation (Aiello, 1993), and an increased tendency to
believe that the quantity of work is more important than its quality (Shell and
Allgcier, 1992). The evidence from several studies is consistent with the social
facilitation prediction that computer monitoring decreases performance on more
complex tasks (Aiello and Svec, 1993) while it can increase performance on very
simple tasks (Aiello and Chomiak, 1992).
One might wonder, in an era in which there is increasing emphasis on build-
ing employee commitment, on working in teams, and on devolving responsibility
as organizations take out layers that make work more complex, how there can be
a concomitant increase in electronic and other surveillance activities—a practice
that research indicates often produces effects opposite those being sought by the
high-commitment work practices so frequently discussed. One avenue to pursue
in understanding this apparent contradiction is to note that, virtually without ex-
ception, the studies of surveillance (and other forms of external control) have
asked about their effects on those subject to these practices. For the practices to
persist, in spite of their often deleterious consequences, it would seem useful to
understand their effects on those administering the practices. It is quite possible
that the control, or at least the illusion of control, produced by implementing
surveillance, and the higher level of involvement and responsibility that accrues
to those who more closely supervise and monitor work, may explain why those
who administer such practices persist in using them even if the results arc not
always positive.
A similar set of issues concerns the use of sanctions. There have now been a
large number of studies of managerial succession showing how various conditions
such as the power of the CEO and the firm's ownership moderate the relationship
between performance and turnover (e.g., Ocasio, 1994; Salancik and Pfeffer,
1980). With few exceptions, these studies have not asked whether or not firing
those associated with poor performance subsequently improves the performance
of the organization. Gavin, Green, and Fairhurst distinguished two strategics of
managerial control, particularly in situations of poor performance: "Managers
may utilize a judicial approach which has its focus on using primarily punitive
measures as a deterrent to future poor performance incidents. . . . Alternatively,
a problem-solving approach may be used which can be characterized as more
interactive, cooperative, and non-punitive in nature" (1995, p. 209).
Gavin et al. (1995), in a laboratory experiment, found partial support for the
ideas that managers use consistent control strategies over time, that they are more
likely to be punitive the worse the level of performance, and that subordinates
perceive punitive control strategies as less just. O'Reilly and Weitz (1980) studied
the behavior of managers in a large retail chain. They found a positive correlation
between the use of sanctions, such as warnings and dismissals, and unit perfor-
mance. They argued that how managers treated poor performers sent an im-
portant signal about the performance culture of the organization. The fact that
firing is often a form of ritual scapegoating (Garrison and Scotch, 1964) is well
known but sometimes overlooked in both theory and empirical research. As Zajac
116 New Directions for Organization Theory
and Westphal (1995) reminded us, all organizational incentives (and sanctions)
have a symbolic, public face to them and can be partly understood in that con-
text.
and four dexterity tests" (p. 21). Then, having scored sufficiently well, she re-
ceived a letter saying she was selected to "enter Phase I of a two phase pre-
employment assessment activity" (pp. 21-22). The assessments, involving working
in groups, were tiring and stressful. It took more than six months to get hired (p.
31):
The reactions of the training class members to the selection process ran from
"grueling" and "demanding" to "fun." . . . Everyone agreed, however, that they
had never worked so hard to land a job. The general sentiment was that we had
really accomplished something and were among a select few. . . . In order to
maintain their status in the selection process, most applicants were forced to
make some degree of sacrifice, (p. 31)
Similar selection processes have been described at New United Motors (Ad-
ler, 1992), even though in this instance, under the United Automobile Workers
contract, the workers had recall rights and the evidence is that even union "trou-
blemakers" were rehired if they applied. Adler noted that "each applicant for the
manager, Group Leader, Team Leader and Team Member positions had to pass
the same three-day assessment test" (p. 120). Initially, 78 percent of the 3,200
hourly applicants were rejected, although most who persisted were later hired.
Why reject those who were later going to be hired? This process tends to create
the feeling of being in a special, highly selected work force. It is also a process
that requires perseverance, and this persistence in efforts to get a job will result
in more commitment to that job. Nor are these commitment tactics used and
useful only for hourly workers. At Data General (Kiddcr, 1981), two project lead-
ers, West and Alsing, used a process called the "signing up" to ensure that those
recruited onto a team of engineers to design a computer had consciously chosen
to undertake a difficult and demanding task and would therefore be willing to
sacrifice themselves and work hard for the project:
"It's gonna be tough," says Alsing. "If we hired you, you'd be working with
a bunch of cynics and egotists and it'd be hard to keep up with them."
"That doesn't scare me," says the recruit.
"There's a lot of fast people in this group," Alsing goes on. "It's gonna be a
real hard job with a lot of long hours. And I mean long hours." (pp. 65—66)
fundamental change in values and behaviors —the goal of socialization —the indi-
vidual must be confronted with evidence that his or her previous social identity
won't bring comfort in the new organizational environment. Hard work, long
hours, and the social isolation that derives from spending so much time at work
all help to make the individual susceptible to social influenees emanating from
within the organization. The stress makes the individual open to these influences.
The corporate hazing also has some of the same effects as other forms of hazing:
it produces a perception that the organization must be worthwhile —otherwise,
one's willingness to suffer so much would be inexplicable. As Aronson and Mills
explained, "Persons who go through a great deal of trouble or pain to attain
something tend to value it more highly than persons who attain the same thing
with a minimum of effort" (1959, p. 177).
The third part of the socialization process involves thorough training through
exposure in the field to the organization's operations and to individuals who ex-
emplify its values. Many retail organizations, such as grocery and department
stores, emphasize putting new employees in the store, regardless of their subse-
quent anticipated career path, to ensure that they fully understand the organiza-
tion and what it is about. Career advancement is slow so that sufficient time is
spent getting thoroughly socialized to the organization's values and behaviors.
The fourth step in the socialization experience entails paying close attention
"to systems measuring operational results and rewarding individual performance.
Systems are comprehensive, consistent, and triangulate particularly on those as-
pects of the business that are tied to competitive success and corporate values"
(Pascalc, 1985, p. 31). What is measured is what is attended to, and measurement
therefore helps to create as well as to reinforce a set of values and norms.
The fifth step in the socialization process involves strict adherence to the
organization's core values. As Pascale (1985, p. 32) noted, subjecting oneself to
an intense socialization experience puts oneself at the mercy of the firm and
leaves one quite vulnerable. In order to convince individuals that socialization
will not work against their interests, the organization must build and maintain a
bond of trust with its employees. That means making promises that ensure indi-
viduals' well-being, and it means keeping those promises. For instance, at New
United Motor Manufacturing, the union and its members gave up the right to
strike over work rules, company contributions to the Supplemental Unemploy-
ment Benefits scheme of General Motors, and the protection of numerous job
classifications in return for higher wages and a promise of no layoffs. When this
no-layoff policy was sorely tested in 1987 and 1988 because of inadequate sales
of the company's cars, the organization found ways of meeting the commitment.
The comptroller and general manager stated, "There's no question in my mind
but that the commitment to no layoffs and the ability of managers to operate on
a day-to-day basis in a way that supports this underlying philosophy are absolutely
crucial to our success. Team members know that when they contribute ideas for
more effective operations they are not jeopardizing anyone's job" (Adler, 1992,
p. 123).
The sixth step in the socialization process is developing reinforcing folklore,
stories, and legends that validate the organization's culture and its aims and that
12O New Directions for Organization Theory
interpret history in ways consistent with its values. Finally, "consistent role models
and consistent traits arc associated with those recognized as on the fast track"
(Pascalc, 1985, p. 33). In other words, through whom it rewards and recognizes,
the organization sends a consistent message about what behaviors and attitudes it
values.
Van Maanen and Schcin (1979) proposed six socialization tactics that could
vary along a bipolar continuum. As reviewed by Ashforth and Saks (1996, p. 150),
these socialization tactics were (1) collective versus individual (were newcomers
grouped together or handled individually?); (2) formal versus informal (were new-
comers separated from the rest of the organization or not?); (3) sequential versus
random (was there a fixed sequence of steps required to assume a new job role?);
(4) fixed versus variable (was there a specified timetable for the socialization pro-
cess?); (5) serial versus disjunctive (was the new member socialized by an experi-
enced member?); and (6) investiture versus divestiture (are the newcomer's iden-
tity and personal characteristics affirmed or stripped away?).
Ashforth and Saks found that, on the one hand, "the six tactics of institution-
alized socialization were negatively related to role ambiguity, role conflict, stress
symptoms, and intentions to quit and positively associated with job satisfaction,
organizational commitment, and organizational identification" (1996, p. 169).
This suggests that more extensive and formalized socialization has positive effects
on those subjected to it. On the other hand, institutionalized socialization was
negatively related "to both attempted and actual role innovation at four months
and to actual role innovation at ten months" (p. 169). But the trade-off between
innovation and change and reduced stress and enhanced job satisfaction may be
more apparent than real. As Ashforth and Saks noted, the issue is what is learned,
not how it is taught, that is important, and it is possible to inculcate norms that
encourage innovation in an institutionalized socialization process.
The evidence suggests that commitment is intentionally built and managed.
There is evidence that the psychological and social psychological processes that
produce commitment are reasonably well understood. There is also literature that
suggests deep ambivalence about commitment, be it to a decision or an organiza-
tion. Being committed is, almost by definition, to some degree a loss of personal
freedom and choice —to be committed to an organization is to stay with it even
in the absence of rational reasons to do so. It is this aspect of commitment that
some find troubling. But, it is important to recognize that individuals often
choose commitment and to be committed. Institutionalized socialization pro-
duced less role stress and higher job satisfaction. Thus, commitment has obvious
positive effects for organizations but also for the individuals who are committed
as well.
Organizational Culture
Although closely related to the idea of control through commitment and socializa-
tion, the topic of organizational culture is somewhat distinct. Commitment is one
important mechanism through which strong cultures may be built (O'Reilly and
Mechanisms of Social Control 121
Chatman, 1996), but the concept of culture incorporates more than simply com-
mitment. Kunda defined culture as:
a learned body of tradition that governs what one needs to know, think and feel
in order to meet the standards of membership. . . . When applied to organiza-
tional settings, culture is generally viewed as the shared rules governing cognitive
and affective aspects of membership in an organi/ation, and the means whereby
they arc shaped and expressed. Of particular concern have been the shared
meanings, assumptions, norms, arid values that govern work-related behavior; the
symbolic, textural, and narrative structures in which they are encoded; and —in
the functionalist tradition —the structural causes and consequences of cultural
forms and their relationship to various measures of organizational effectiveness.
(1992, p. 8)
The definition of the concept of culture has occasionally been contested and
lacks consensus. Some (Reichcrs and Schneider, 1990) have argued that culture
is conceptually not distinguishable from the construct of organizational climate.
Culture has been defined and measured at different levels of analysis, including
the group, the entire organization, or even the industry (e.g., Chatman and Jehn,
1994). O'Reilly and Chatman have noted that some "define culture as what an
organization is while still others argue that it is what an organization has (Schein,
1985; Smircich, 1983)" (1996, p. 159). O'Reilly and Chatman offered a definition
that is grounded in the view of culture as a form of organizational control: "we
define culture as a system of shared values (that define what is important) and
norms that define appropriate attitudes and behaviors for organizational members
(how to feel and behave)" (p. 160). Defined in this way, one can ask, empirically,
to what extent there is agreement about values (consistency) and to what extent
violations of organizational norms and rules are sanctioned severely (the intensity
with which values and beliefs about appropriate behavior are held). This defini-
tion is, therefore, quite amenable to empirical, tjuantitative measurement and,
moreover, links culture with other social psychological constructs such as norms
and values.
Given the disagreements about the definition of culture, it is not surprising
that there is also a variety of opinions about how to operational ize and measure
the concept. One approach, the Organizational Culture Profile (O'Reilly, Chat-
man, and Caldwell, 1991), employs a Q-sort profile comparison process that asks
respondents to put fifty-four value statements into a set of nine categories from
"most characteristic of my firm's culture" to "most uncharacteristic of my firm's
culture." The sorting process permits fewer cards at the extremes than in the
122 New Directions for Organization Theory
middle. The procedure produces a measure with good test-rctcst reliability. With
an average of thirty-five respondents from fifteen firms in various service indus-
tries, Chatman and Jehn (1994, p. 531) observed a high degree of consensus
within firms, although the extent of consensus varied, as might be expected as
some firms had stronger cultures than others. In several studies using the OCP, a
consistent set of seven factors emerged: "(1) innovation, (2) stability, (3) an orien-
tation toward people, (4) an orientation toward outcomes or results, (5) an em-
phasis on being easygoing, (6) attention to detail, and (7) a collaborative or team
orientation" (p. 534). Chatman and Jehn demonstrated that even within the ser-
vice sector, individual firms had unique cultures. They also found that "industry
differences explain more variance than organization differences for six of the
seven culture dimensions" (p. 537). It seems fair to state that progress has been
made in operationalizing the concept of culture and developing measurement
procedures that produce results consistent with theoretical expectations and that
are replicable. As noted below, however, fragmentation in the culture research
literature, compared to the literature on personality and organizational behavior,
for instance, has resulted in less borrowing of measurement processes across stud-
ies and consequently less cumulation of results.
Since the mid-1980s there have been a plethora of both scholarly and
practitioner-oriented books on the subject of culture. Barley, Meyer, and Gash
(1988, p. 33), searching six bibliographic data bases, noted that the total number
of papers published annually on organizational culture increased from less than
twenty in 1979 to about fifty in 1984 and to close to 130 in 1985, an exponential
rate of growth. Some of this activity has been stimulated by the idea that strong
cultures arc positively related to economic performance (e.g., Kottcr and Heskett,
1992; Dennison, 1990), that managing culture is possible (e.g., S. Davis, 1984;
O'Reilly, 1989), and that culture is an effective managerial method used in exem-
plary organi/ations such as Hewlett-Packard (e.g., Deal and Kennedy, 1982). Bar-
ley ct al. also noted that "the upsurge of interest in organizational culture in large
part reflected the back-to-back commercial success of three bestsellers that spoke
of culture under various guises. . . . By attracting attention in both the academic
and managerial communities, organizational culture attained the status of a domi-
nant idea in a relatively short period of time" (1988, p. 32).
Not all empirical studies have found a relationship between strong cultures
and performance (e.g., Sichl and Martin, 1990), and even those that have uncov-
ered an effect have noted the relationship is often complex. Thus, for instance,
Kotter and Heskett found that in order to enhance performance, culture needed
to be matched to the firm's strategy, and even then, "strategically appropriate
cultures will not promote excellent performance over long periods unless they
contain norms and values that help firms adapt to a changing environment"
(1992, p. 142).
Unfortunately, the vast amount of writing on the subject of organizational
culture has been accompanied by a fragmentation that makes this topic domain
an excellent exemplar of the costs of an absence of paradigmatic consensus.
There is agreement on the fact that "research on organizational cultures is highly
fragmented. Empirical research addresses a wide variety of content themes and
Mechanisms of Social Control 123
Cultural Ideal-types
The first thing you do is teach workers the techniques of work analysis. Next . . .
you get the workers as a group to time each other with a stopwatch. . . . After
everybody has been timed, workers analyze what they think is the best perfor-
mance and break that process down into little pieces. . . . All this leads to a way
of doing the job that everybody agrees with. . . . The point is to get workers to
participate in defining the standards and encourage them to constantly make
suggestions to improve them. (Adler, 1992, p. 141)
Everyone, from the company president on down, wore the same basic uniform,
parked in the same general parking lots, ate in the same cafeteria, and was re-
stricted to a half-hour lunch period. . . . Clerical and management Associates
worked together in one large room without barriers. . . . SIA used rituals and
ceremonies to create a shared experience of belonging. These included morning
exercises and daily team meetings, (p. 110)
126 New Directions for Organization Theory
Even though the equality and egalitarianism was not always authentic, the steps
taken to provide evidence of these values did affect how employees felt about and
reacted to their work environment.
Third, cultures arc created through shared social information — informational
social influence (Dcutsch and Gerard, 1955). Particularly when people first enter
an organization, they arc often uncertain about what to do and what the rules of
the game are. If they arc exposed to consistent social information —consistency
in norms— about expectations for behavior, they are likely to conform to those
expectations. Consistency is often difficult to maintain under immediate perfor-
mance pressures but is critical to building a strong culture and implementing
cultural control, as Adler's reporting of the observations of a NUMMI manager
indicated: "At GM, it's easy to slip into the mentality of 'just do it—I've got a
production schedule to meet.' The biggest challenge for managers coming into
NUMMI is the absolute commitment to consistency to all our principles, not just
to a production schedule set by Marketing. At NUMMI, we've got to walk like
we talk" (p. 127). Many strong-culture organizations attempt to manage the infor-
mational environment by isolating, either physically or socially, through hard
work and inclusive social activities, individuals from other social influences that
would be distracting or diversionary, such as other organizations or outside inter-
ests.
Finally, organizations shape behavior through reward systems, although such
systems frequently don't use monetary rewards. Biggart's (1989) study of direct
sales organizations, such as Mary Kay and Tupperware, found that such organiza-
tions used frequent meetings, recognition of achievements, awards, and even visits
from a charismatic founder to reinforce behavior consistent with cultural values.
Underlying all of these forms of cultural control is always the implicit threat that
those who don't fit in will be expelled from the organization, and this fear of
social ostracism is also a strong factor promoting conformity to the organization's
core values and decision premises.
Leadership
Leadership Effects
Johns noted that organizations might do things that restrict variation in leader
consideration and that subordinate performance could he affected by equipment
or other conditions. He noted that "constraints may exist that restrict required
variance and thus threaten the hypothesis in spite of an actual true-score relation-
ship between consideration and performance in the population" (p. 81). These
comments on the effects of constraints on evaluating hypotheses are applicable to
literatures other than that on leadership, but the arguments are particularly im-
portant in this context because of the very real selection and behavior constraints
often encountered. These constraints at once reduce the likelihood of finding
leadership effects and also make such effects less likely to exist.
Another variant of the argument to use caution in assessing the effects of
leaders is that one needs to be careful in inferring effects of leader behavior in
natural situations, because it is quite likely that managerial style itself adjusts
to differences in performance. An observed relationship between leadership and
performance might well be because performance causes leader behavior rather
than the reverse (Lowin and Craig, 1968).
The debate about whether or not leadership matters has occurred on two
fronts—theoretical and empirical. The empirical issue was joined with a study by
Lieberson and O'Connor (1972) that decomposed the variation in sales, profits,
and profit margins (profits/sales) for a sample of 167 large corporations over a
twenty-year period to effects associated with year, industry, company, and leader-
ship (as represented by the chief executive). The study found, not surprisingly,
that there were much larger industry and company effects than leadership effects.
Salancik and Pfeffer (1977) replicated this methodology in an examination of the
effect of mayors on city budgets, specifically spending on different categories.
Again, there was only a small effect of leadership observed. The methodology
employed in these studies depends importantly on how the variance is decom-
posed (the order in which factors are considered) and also requires one to esti-
mate a time lag (including no lag, or contemporaneous effects) between leader-
ship succession and a change in the variable being studied. Moreover, effects
were described in relative terms. In absolute terms, a small percentage change in
either sales, profits, or budgetary expenditures can amount to a lot of actual dol-
lars, given the size of the units that were examined.
Weincr and Mahoney (1981), using a somewhat different methodology that
included more direct measures of company, industry, and period effects (e.g.,
organizational size, industry sales, and capital structure), observed substantially
larger effects for leadership in their study of 193 business firms over a nineteen-
year period. However, neither of their measures of leadership strategy explained
much variance. They argued that "means to identify causal variables of leadership
at the corporate level are needed" (p. 469). Meyer's (1975) study of government
finance offices from 1966 to 1972 discovered that there were more changes in
organizational structure in offices that selected new leaders compared to finance
departments in which there were no changes in leadership.
Two studies illustrate different methodologies that are more useful and valid
in assessing leadership effects. Lieberman, I.au, and Williams (1990) used Cobb-
Mechanisms of Social Control 129
concept has thereby gained a brilliance that exceeds the limits of normal scien-
tific inquiry" (1985, p. 78).
Meincll et al. (1985) introduced the idea of the romance of leadership, an
elevated belief in the efficacy of leaders and leader behavior. They found an
inverse relationship between the relative number of doctoral dissertations con-
cerning leadership and economic conditions, indicating that there was more inter-
est in leadership when economic conditions were worse. They also observed a
positive relationship between the number of articles published in the Wall Street
Journal concerning leadership and firm performance and, in a set of experimental
studies, found thai extreme performance outcomes were more likely to result in
attributions of leadership effect. In an analysis of their data that explored the
extent to which there is a desire to attribute performance to human agency, re-
gardless of hierarchical position, Meincll et al. observed statistically significant
positive correlations between attributions to leaders and to subordinates in the
experimental studies. The idea of the romance of leadership is persuasive, but
again, there has been insufficient pursuit of an initially promising idea. Two direc-
tions would seem to warrant exploration. First, it is highly unlikely that people
are born with the idea that leaders matter. The romance of leadership is learned,
and that stimulates the question of how and under what conditions some come
to believe in leadership effects more than others. It is likely that the degree to
which people have a romance of leadership varies across cultures, with possibly
more relevance in more individualistic cultures. Second, it would be worthwhile
to explore factors other than extreme performance or performance problems that
produce an increased attribution of leadership effects.
Even if one accepts both the empirical and theoretical critique of the leader-
ship literature (and certainly not everyone docs), the focus on leadership is quite
understandable and rational, given a pragmatic orientation. Even if a leader's
effects are constrained, someone in a leadership position must still do something.
After all, one cannot affect general economic conditions or the history of an
organization. Decisions about whom to choose for key positions remain, while
occupants of those positions have to take some action, regardless of how effica-
cious that action may or may not be. In that sense, the arguments about leader-
ship effects are theoretically interesting but possibly managerially and practically
irrelevant—people must be chosen for positions, trained to he as effective as they
can be in them, and, once in the jobs, do something, whether the impact is large
or small. It therefore remains important to develop research that can at least help
in addressing these three issues.
Nevertheless, the comments of Meindl and his colleagues are useful in help-
ing us understand how the leadership literature has evolved. For instance, the
current emphasis on building organizational vision and visionary organizations
(Collins and Porras, 1994) and the relative neglect of leader behaviors such as
consideration and initiating structure, behaviors that used to be studied (Fleish-
man, 1973) —arc comprehensible given the less hierarchical nature of many of
today's organizations. This has made the exercise of hierarchical, role-based lead-
ership less relevant and the task of building the ability to take coordinated collec-
tive action in the absence of direct hierarchical authority more important.
Mechanisms of Social Control 131
On the question of what effective leaders do, the leadership literature has been
particularly interested in distinguishing charismatic or transformational leaders
from transactional leaders. Bass (1985) argued that transactional leaders exchange
rewards for behavior provided by followers while transformational leaders develop
relationships with followers that inspire and stimulate them to forgo their own
self-interest for some higher collective good. Shamir, House, and Arthur (1993)
argued that noncharismatic leaders affected followers' cognitions, such as their
calculations about the desirability of engaging in some behavior, and possibly
their ability to perform specific tasks. By contrast, charismatic leaders affected
followers' emotions and self-esteem or self-concept. Shamir et al. operationalized
charismatic leadership as an interaction and communication process between a
leader and others that results in the internalization of the leader's goals and val-
ues, a strong moral or personal (as contrasted with a calculative) commitment to
those values and goals, and a tendency on the part of the followers to submerge
their own interests and identity for the sake of the collectivity. House (1977) has
noted that one charismatic behavior is articulating ideological goals, which he
maintained had a more potent effect on the motivation and self-confidence of
followers than more utilitarian or pragmatic objectives. Shamir et al. (1993) noted
that, in comparison to noncharismatic leaders, charismatic leaders would tend to
use more references to values and moral justifications, make more references to
the collective, use more references to history and to more distant objectives, make
more positive statements about followers' self-worth, and communicate higher
expectations for the followers.
House (1977) argued that charismatic leaders differed from other leaders not
only in their behaviors but also in their own personal characteristics. In particular,
he linked the study of dispositions to the study of leadership by arguing that
leaders that have charismatic effects on subordinates differed from those that did
not have such effects in terms of their dominance, self-confidence, need for in-
fluence or power, and strong belief in the righteousness of their ideas.
Charismatic leadership is not likely to be exhibited equally in all circum-
stances. Shamir et al. (1993) argued that charismatic leadership is more likely to
emerge under conditions of greater uncertainty and ambiguity—when perfor-
mance cannot be easily quantified or measured, when means to achieve objec-
tives arc uncertain, and when it is difficult to link extrinsic rewards to individual
achievement. They also noted that charismatic leadership is more likely under
exceptional organizational conditions such as rapid changes in the environment,
performance pressures, reorganizations, mergers, or sudden growth. Note that the
conditions for the emergence of charismatic leadership are similar to those that
Meindl et al. (1985) argued were associated with the romance of leadership. But
that link is logical. The conditions that lead observers to attribute leadership and
to want to feel control are precisely those conditions that would make the appear-
ance of charisma more probable.
Because of the issues of attribution already discussed, experimental studies of
charisma offer the cleanest exploration of these ideas. Howell and Frost (1989)
132 New Directions for Organization Theory
language in the task of leadership. Eccles and Nohria noted, "To view manage-
ment from a rhetorical perspective is to recognize that the way people talk about
the world has everything to do with the way the world is ultimately understood and
acted in. . . . The primary task of managerial language has always been to per-
suade individuals to put forth their best efforts in a collective enterprise with other
men and women" (1992, p. 29). Consequently, "the basic task of management
. . . is to mobilize action by using language creatively to appeal to the self- and
collective identities of individuals" (p. 37).
The symbolic, rhetorical role of leadership would seem to rely on linguistic
skills, having a good vocabulary, and possessing the ability to select images and
language that reach and touch others. There is some case evidence that this is
true. Eccles and Nohria (1992) compared the language of Jack Welch, CEO of
General Electric and an individual who led it through an important transforma-
tion, with that of Roger Smith, the CEO of General Motors, who is generally
acknowledged not to have helped the organization become better at competing
in the automobile industry. They describe Welch's language as conveying "an
imaginative vision of the future, a realistic portrayal of the present, and a selective
depiction of the past which can serve as a contrast to the future" (p. 32). By
contrast, Smith "externalized GM's problems" and never provided enough detail
to his vision of building a twenty-first-century corporation to inspire action. Rheto-
ric that is effective in mobilizing action also maintains a balance between clarity
and ambiguity:
effective rhetoric aims to be clear but never too clear. It aims to be robust across
as many different situations as possible, and to be flexible enough to incorporate
the different meanings, emphases, and interpretations that different people will
inevitably give to it. . . . The concept of robustness also implies that the rhetori-
cal strategy should be dynamic, adjusting over time to evolving contingencies.
(p. 35)
To the extent rhetorical skill is critical for leaders, leadership training might
profitably involve training in drama and literature. It would be interesting to as-
sess the extent of such training in conventional leadership education as well as to
evaluate the effectiveness of this type of skill building compared to more cognitive
and rationality-based leadership development. The literature on management as
symbolic action presumes that there is enough uncertainty or ambiguity in situa-
tions such that language can direct and divert attention, creating meaning out of
ambiguity. This symbolic role of management has been emphasized in writings
about organizational culture (e.g., O'Reilly and Chatman, 1996) and is another
link between these mechanisms of social control.
boom industry in its own right. There have been literally decades of study about
leader behaviors and their effects, and even the elements of charisma have been
studied. Consequently, one might expect to see a large literature on whether or
not leadership training makes a difference and which approaches are the most
effective. However, such a literature is virtually nonexistent. As Gaziel noted, "it
is ... difficult to find reports on empirical investigations concerning the degree
to which academic studies of management develop the skills and qualifications in
their students required for successful performance of managerial roles" (1994, p.
341), and a similar scarcity of reliable research exists for other forms of manage-
ment and leadership training as well. There are studies that rely on the self-reports
of those who have experienced leadership training. Such reports should, of
course, be viewed with a great deal of caution. Having invested time and energy
in a course of study, one is likely to rationalize that investment by saying that it
was helpful and useful. Moreover, many of these courses are at a minimum more
pleasant than work and often allow students to bond with the other members of
the class, and each of these effects would also be expected to produce high levels
of participant-reported satisfaction with the programs.
Kirkpatrick (1979) has developed a useful overview of how to evaluate the
effectiveness of leadership or, for that matter, any other kind of training. He ar-
gued that four criteria should be employed: (1) attitudes and perceptions of those
who participated in the program; (2) learning, which refers to mastery of princi-
ples, concepts, facts, and techniques; (3) behavior, or whether participants actu-
ally do things differently as a consequence of having been exposed to the training;
and (4) results, which refers to whether or not morale and turnover improves,
whether economic performance is better, and so forth. Of course, doing a com-
pletely valid study of these four dimensions would be difficult, as the problems of
inference are severe even with a control group design. Nevertheless, it does seem
that given the substantial investment of effort, it would be useful to know whether
the effort is worthwhile and, perhaps even more important, what forms of training
and education arc comparatively more efficacious under what circumstances.
In a study of the use of an assessment center process for developmental pur-
poses, Engelbrecht and Fischer (1995) did find that the experience did have an
effect even three months after participants had gone through it. Compared to a
control group, those managers who went through the developmental assessment
center process differed "in action orientation, task structuring, development, em-
pathy, managing information, and probing. No significant difference was found
in either synthesis or judgment. The results . . . indicate an improvement in
managerial performance regarding the six ... dimensions" (p. 397). Using be-
haviorally anchored rating scales, the study indicated that "the developmental
assessment center process can . . . be considered to be accountable for 14.3% of
the total variance in overall managerial performance . . . 3 months after center
attendance" (p. 398). This study provides evidence that interventions can affect
managerial behavior. It would be wonderful to have many more studies of this
type, with samples large enough to indicate under what conditions these interven-
tions have the greatest effect.
Mechanisms of Social Control i^
Conclusion
the biases and problems that beset negotiations have drawn increasing scholarly
attention.
Power has had a bad name in social science research and is most often conspicu-
ous by its absence from the literature. It is not just Williamson (e.g., Williamson
and Ouehi, 1981) who has critiqued the concept and argued that it was neither
necessary nor useful for understanding organizations. Courses about power and
influence in schools of administration started only in the late 1970s and still exist
in comparatively few institutions. A perusal of management texts at that time and
even later reveals that power was often omitted entirely from consideration. When
it was discussed, the emphasis was on a relatively simple conception of interper-
sonal power—French and Raven's (1968) bases of power (expert, reward, sanc-
tion, and so forth). Perrow noted that "the literature on power in organizations is
generally . . . preoccupied with interpersonal or intergroup phenomena . . . or
else it takes as the major dimension of power the relative and absolute power of
levels in the hierarchy" (1970a, p. 60). Kipnis, Schmidt, and Wilkinson wrote that
"organizational psychologists have not been particularly interested in studying the
ways in which people at work influence their colleagues and superiors to obtain
personal benefits or to satisfy organizational goals" (1980, p. 440). Although power
may have always been important and certainly is becoming more so as hierarchi-
cal bases of authority erode under the environmental changes described above,
power as an idea docs not fit very well with either our conception of organizations
and individuals as being rational utility and efficiency maximizers or with our
social values that emphasize cooperation and deernphasize inequality in the ac-
cess to social resources.
We are profoundly ambivalent about power (Pfeffer, 1992) and that ambiva-
lence has led to recurrent questioning of the concept and its definition. Illustrat-
ing the pervasiveness of the ambivalence, Kanter noted, "Power is America's last
dirty word. It is easier to talk about money—and much easier to talk about sex—
than it is to talk about power" (1979, p. 65). John Gardner wrote, "In this country
. . . power has such a bad name that many good people persuade themselves
they want nothing to do with it" (1990, p. 55). Gandz and Murray, from a survey
of 428 managers, provided empirical evidence of the ambivalence: although the
managers overwhelmingly agreed with the statements "the existence of workplace
politics is common to most organizations" and "successful executives must be
good politicians," they also thought that "organizations free of politics are happier
than those where there are a lot of politics" and "politics in organizations arc
detrimental to efficiency" (1980, p. 244).
After first considering the definition of power and issues of its measurement,
we address three fundamental questions concerning power in organizations: (1)
under what circumstances are power and influence more important in organiza-
tional decision making, and under what circumstances are political skills less
critical for getting things done? (2) what are the important sources of power? and
138 New Directions for Organization Theory
(3) what arc the strategics and tactics through which interpersonal influence is
developed and exercised and power gets employed?
Suppose we imagine that each power encounter occurs between just two people
chosen al random from the total population. . . . Further, assume that at each
encounter we will decide who prevails by flipping a coin. If the total number of
Developing and Exercising Power and Influence 139
encounters per person is relatively small and the total number of persons rela-
tively large, such a process will yield a few people who are successful in their
encounters virtually all of the time, others who are successful most of the time,
and so on. . . .All of the chance models generate power distributions. They are
spurious distributions in the sense that power, as we usually mean it, had nothing
to do with what happened. But we can still apply our measures of power to the
systems involved. (1966, pp. 51-52)
March noted that to distinguish power from chance, one needed to assess (1)
whether power was stable over time, (2) whether power was stable over subject
matter, (3) whether power was correlated with other personal attributes, and (4)
whether power was susceptible to experimental manipulation. He argued that
there was evidence for the presence of all of these conditions, and since that time
in particular an extensive experimental literature on power (e.g., Cook et al.,
1983) has emerged.
Nevertheless, March was concerned about the overuse of the concept of
power in analyzing social decision situations. Because power was so ubiquitous
and obvious in conversations about everyday life, there is a tendency "to assume
that it is real and meaningful. . . . However . . . we run the risk of treating the
social validation of power as more compelling than it: is simply because the social
conditioning to a simple force model is so pervasive" (1966, pp. 68-69). He also
was concerned that because of the obviousness of power, it would come to be
used as a residual category for explanation—a name for the unobserved variance
in research studies. As Dahl noted, "a Thing to which people attach many labels
with subtly or grossly different meanings in many different cultures and times is
probably not a Thing at all but many Things" (1957, p. 201). Unless considered
and studied quite carefully, the force activation model of power is quite difficult
to disprove:
It is clear from a consideration both of the formal properties of activation models
and of the problems observers have had with such models that they suffer from
their excessive a posteriori explanatory power. If we observe that power exists and
is stable and if we observe that sometimes weak people seem to triumph over
strong people, we are tempted to rely on an activation hypothesis to explain the
discrepancy. (March, 1966, p. 61)
In studies of power, it is essential to measure power separate from its use and
results and to define and measure power before the analysis of the social choice
rather than after. Again, however, since these initial concerns, a number of studies
have been reasonably successful in overcoming these empirical and theoretical
difficulties. Nonetheless, it is wise to keep them in mind in evaluating the litera-
ture on power.
There is an alternative perspective that maintains that the very ability of re-
spondents to so quickly identify power and to discuss it provides evidence of the
validity of the concept. Perrow has argued forcefully for this position:
Power is a preoccupation of the managers in the firms, as evidenced by the
interview data. . . . Given the nature of the concept and the "reality," I am not
disturbed that the term "power" has different meanings. . . . A single, consistent
140 New Directions for Organization Theory
profit boards a manager sat on; (3) the financial standing of the corporate boards
of which an executive was a member; and (4) whether or not the manager had
an elite education.
Finkelstein tested his measures of power on a sample of 1,763 top managers
from 102 firms in three industries over a five-year period. A factor analysis of all
of the indicators reproduced the four scales. Cronbach's alphas for each of the
scales were reasonably high, ranging from .67 for prestige power to .83 for struc-
tural power. A second study that asked for the perceived power of individuals
revealed that perceived power was most strongly correlated with structural power
and prestige power and was not statistically significantly correlated with expert
power and was only weakly correlated with ownership power. Finally, Finkelstein
reported that the consideration of the power of senior managers permitted better
predictions of diversification strategics than models that did not include the power
indicators.
Schein's article did not provide any evidence for the claim that efficiency
considerations would drive out at least some forms of power behavior—the most
self-serving—and this remains an important empirical question. There is the
problem with this argument that the very definition of how much slack or how
much environmental pressure there is, the question of what is the most appro-
priate response, and the issue of which individuals or units are best suited to
coping with the problem are all settled at least in part through a political process.
Surveys of managers by Gandz and Murray (1980) and Madison et al. (1980)
asking what levels, functions, and decisions in organizations seemed to most in-
volve the use of power revealed that power was used more in major decisions,
such as reorganizations or personnel changes, and was used more at senior organi-
zational levels, which also presumably deal with more important decisions. Their
results also indicated that power was used more in areas in which there was more
uncertainty and in which performance was harder to assess (e.g., staff rather than
line production operations).
This idea that power would be used more under conditions of uncertainty or
ambiguity has been examined in a number of studies focusing on science and
academia. The basic idea came from a typology of decision situations developed
by Thompson and Tuden (1959). They categorized decision situations according
to whether or not there was agreement about goals or preferences and whether or
not there was agreement about technology—the connections between actions and
their consequences. A summary of their typology noted:
computational decision making procedures involved in rational choice are used
in decision making only when there is agreement both on the goals and on the
connections between actions and outcomes. When there is no agreement on
goals, compromise is required to reach a decision. When there is no agreement
on technology, judgment is necessary to determine the best course in achieving
consensually shared goals. And when there is neither agreement on goals nor on
technology, an unstructured, highly politicized form of decision making is likely
to occur. (Pfeffer, 1981, p. 71)
Sources of Power
Research has identified numerous sources of power for both individuals and orga-
nizational subunits. Some sources of power are personal characteristics. R. W.
Allen et al. (1979) interviewed eighty-seven managers, including thirty chief exec-
utives, in southern California electronics firms to assess their beliefs about the
personal characteristics associated with being powerful and being effective in the
use of organizational politics. The characteristics mentioned most frequently
were: articulate, sensitive, socially adept, competent, and popular. The study-
raises two issues. First, in relating personal characteristics to power, one needs to
be careful that the characteristics associated with power (for example, self-
confidence or being popular) are not a consequence of power rather than its
source. It is reasonable to argue that one will be more socially adept and popular
to the extent one has power—these characteristics may be both products and
sources of power. The characteristics may also not be independent of each
other—articulate and sensitive people may be more popular as a result. Second,
there is the question of the extent to which these characteristics arc relatively
fixed or are malleable. Some, such as sensitivity and articulateness, can presum-
ably be affected by training in clinical skills and public speaking, respectively.
Other sources of power are structural, deriving from one's position in the
division of labor or one's location in the network of communication. As Perrow
noted, "the preoccupation with interpersonal power has led us to neglect one of
the most obvious aspects of this subject: in complex organizations, tasks are di-
vided up between a few major departments or subunits, and all of these subunits
are not likely to be equally powerful" (1970a, p. 59). One structural source of
power is the ability to cope with the uncertainty an organization faces (Hickson
et al., 1971; Minings et al., 1974). Since the form and amount of uncertainty can
vary, depending on the particular decision and situation, so too do the characteris-
tics associated with having power.
Developing and Exercising Power and Influence 145
lished at the time of a firm's founding tended to persist in his sample of semicon-
ductor firms. Studying subunits in health care clinics, Lachman (1989) observed
that previous power position was the best predictor of subsequent power. One way
in which power is self-perpetuating is through the impact of the effective use of
power on an individual's reputation, which then becomes a source of power:
"Most observers would agree that present reputations for power are at least in part
a function of the results of past encounters. . . . Most observers would probably
also agree that power reputation, in turn, affects the results of encounters" (J. G.
March, 1966, p. 57).
Power also derives from formal position or authority. Brass and Burkhardt
(1993) reported that level in the organizational hierarchy was significantly related
to others' perceptions of a focal individual's power. Sometimes that formal posi-
tion can be used to protect the powerholder from the consequences of poor per-
formance. Bocker (1992) studied sixty-seven organizations over a twenty-two-year
period and asked what happened to chief executives when organizational perfor-
mance was poor. He found that powerful chief executives were able to survive
performance downturns by displacing the blame onto their subordinates, who
tended to be replaced when perfonnance suffered. By contrast, less powerful chief
executives, who had less power because they owned less of the company's stock
and there were more outsiders on the board of directors, were not able to displace
poor perfonnance and fire their subordinates; rather, they were replaced them-
selves.
Power is built by doing favors for others. Studies of the power of chief execu-
tives and its effects consistently find that the higher the proportion of outside
directors the CF,O has appointed, the more favorably that individual is treated.
The argument is that being appointed to the board by someone activates a norm
of reciprocity, and one is tied to the individual who did the appointing. Wade,
O'Reilly, and Chandratat (1990) noted that there was a higher incidence of the
provision of golden parachutes —contracts that pay off executives if they lose their
jobs because of a change in control of the corporation, through a takeover, for
instance — for CEOs who had been able to appoint more outsiders to their boards.
Schwartz noted that waiting is a form of investment and, like other forms of
investment and commitment, increases the value of the object being sought. If
one did not think the person waited for was important or valuable, then one
would have engaged in a senseless activity. Conversely, the more one waits, the
more important and valuable must be that which is waited for, to make sense of
one's own behavior:
if we regard waiting for a scarce service as an investment or sacrifice in return
for a gain, we may measure part of the value of the gain by assessing the degree
of sacrifice occasioned on its behalf. . . . The subjective value of the gain is
therefore given not only by the objective value of the service but also by the
amount of time invested in its attainment. . . . The other's service becomes
valuable (and he becomes powerful) precisely because he is waited for. (1974, p.
857)
These initial insights on the relationship between time, waiting, and power
have not been extensively pursued in subsequent empirical study. This seems a
shame, because, as Schwartz noted, "queuing for resources is ... a fundamental
practice of social organization" (1974, p. 842). Much could be learned about
power distributions by examining patterns of waiting time, and the use of delay
or waiting to create or ratify power remains to be empirically explored, particu-
larly as to the boundaries beyond which this strategy may backfire.
In an attempt to systematically understand influence tactics, Kipnis, Schmidt,
and Wilkinson (1980) developed a fifty-eight-item questionnaire focused on vari-
ous targets of influence (peers, superiors, and subordinates) that was factor ana-
lyzed to extract eight dimensions of influence: assertiveness, ingratiation, rational-
ity, sanctions, exchange, upward appeals, blocking, and coalitions. They found
that the use of these influence tactics varied systematically with the context in
which they were attempted. The higher the status of the target person, the greater
the reliance on rationality tactics (e.g., writing a memo, using logic, or making a
plan). Assertiveness was used more with lower-status targets, and exchange was
used most frequently with co-workers. The reasons for exercising influence also
affected the choice of tactics: personal assistance prompted the use of ingratiation,
while efforts to improve a target individual's performance tended to rely on assert-
iveness and rationality. Kipnis ct al. found that "respondents showed the least
148 New Directions for Organization 1'heory
Their focus was on Senator Gurncy's attempts to discredit the very damaging
testimony of John Dean. Gurney was a Republican and a partisan defender of
President Nixon; John Dean had been the president's legal counsel and provided
evidence concerning Nixon's role in the Watergate cover-up. Gurney's strategy
was to demand the literal truth, free of any interpretive context—to demand just
the facts. When Dean could not provide this "literal truth" and sought to explain
the circumstances of a particular conversation, Gurney used this as a way to
discredit Dean's account. As Molotch and Boden noted, "if all accounts are poten-
tially vulnerable to the challenge that they are not really objective, then any conver-
sationalist is open—at all times —to the charge of interactive incompetence"
(p. 274, emphasis in original). They concluded by noting that "demands for 'just
the facts,' the simple answers, the forced-choice response, preclude the 'whole
story' that contains another's truth" (p. 285). Gurney was able to translate his
potential power advantage, derived from being a U.S. senator, the questioner in
the proceedings, and Dean's admission of personal culpability in the events, into
an interactional advantage through the use of a conversational ploy of asking only
for the literal truth. In addition to providing a specific example of a conversational
strategy used for exercising power, the Molotch and Boden study illustrates the
potential for this line of analysis to illuminate other situations in which talk is
part and parcel of how power is accomplished.
Emotion is, along with language, another important way in which influence
is exercised in organizational settings. As Rafaeli and Sutton wrote, "The view
that organization members routinely use expressed positive emotions as tools of
social influence is ubiquitous in organizational behavior" (1991, p. 749). Ex-
pressed negative emotion can also be a potent influenccr. Sutton's (1991) study
of bill collectors found that negative emotions could be part of an influence
strategy of negative reinforcement—when the debtor paid (or at least promised to
pay), the anger and disapproval expressed by the bill collector stopped. Thus, the
debtor was reinforced (a noxious external stimulus was removed) for agreeing to
pay. "Bill collectors believe that expressed emotions such as anger, irritation, and
mild disapproval serve as tools of social influence when such conveyed feelings
induce anxiety in debtors, and debtors construe that escape from such anxiety
will be the consequence of complying with the collectors' demands" (Rafaeli and
Sutton, 1991, p. 750).
In an ethnographic study of bill collectors and Israeli police interrogators,
Rafaeli and Sutton explored the use of emotional contrast strategies as tools of
social influence. The principle is that "stimuli that are presented before, during,
or after a given stimulus shape its meaning" (1991, p. 750), so that what is nice
or noxious behavior is very much affected by the context in which that behavior
150 New Directions for Organization Theory
occurs. Thus, for instance, Kipnis and Vandervccr (1971) found that subjects in
the role of supervisor were more likely to give larger pay raises and higher perfor-
mance evaluations to subordinates attempting influence through ingratiation
when a hostile co-worker was present as contrasted to situations in which that
hostile co-worker was not present. The Rafaeli and Button data suggested that five
emotional contrast strategies were employed: "(1) sequential good cop, bad cop,
(2) simultaneous good cop, bad cop, (3) one person playing both roles, (4) good
cop in contrast to hypothetical bad cop, and (5) good cop in contrast to expecta-
tions of bad cop" (1991, p. 758). They argued that the presence (or threatened
presence) of both styles of behavior made the good cop behavior appear nicer
and the bad cop behavior appear nastier than if either were presented alone.
They concluded that there were three mechanisms through which the perceived
amplification of expressed emotion, accomplished by using emotional contrast
strategies, worked:
First, target persons may experience accentuated anxiety in response to bad cops
and accentuated relief in response to good cops when there is a contrast. . . .
They may acquiesce with requests for compliance in order to escape from the
anxiety or fear. . . . Second, the contrast may accentuate the targets' perceptions
that good cops are kind arid helpful. As a result, the targets may feel pressure to
reciprocate the kindness by complying with the good cops' wishes. Third, accen-
tuated feeling of relief in response to good cops . . . may lead target people to
develop trust in the good cops. (pp. 764-766)
Our understanding of power and influence processes has advanced substan-
tially through research using multiple methods. What that work has not done yet
is to explore the scope conditions for the various influence strategies and sources
of power. We know less than we might about how strategies of social influence
fail, about when theoretically predicted determinants of power don't predict actual
power, and about when power is used in situations either more or less than pre-
dicted by the context. Consequently, some important research remains to be
done.
Negotiation in Organizations
Up until the 1990s, virtually all of the literature on negotiation ignored con-
text almost completely, focusing either on the structure of payoffs and relative
power or on individual differences in negotiation behavior. More recently, much
of the research has focused on cognitive biases because of the premise that "to
negotiate most effectively negotiators need to make more rational decisions"
(Neale and Bazerman, 1991, p. 1). This research builds on the extensive work on
biases in decision making (e.g., Tversky and Kahneman, 1974) under the premise
that "negotiation represents a special case of decision making" (Neale and North-
craft, 1990, p. 56). This cognitive emphasis contrasts with the game theoretic
approach to analyzing strategic interaction in economics, an approach that im-
plies that "the structure of the negotiation determines its outcomes. The behavior
and cognitions of players during negotiations simply represent the unfolding . . .
of the structure. . . . Similarly, game theory allows no explicit role for the social
processes that behavioral scientists often argue characterize negotiation (Neale
and Bazerman, 1991, p. 9).
The decision- or cognitive-bias approach also contrasts with an earlier focus
on the effect of individual differences and personality on negotiation processes
and outcomes. Neale and Bazerman, reviewing the extensive literature on the
effect of individual differences such as gender, cognitive complexity, locus of con-
trol, self-esteem, and risk preferences on negotiations, concluded that "individual
differences offer little insight into predicting negotiator behavior and negotiation
outcomes" (1991, p. 4). These authors also eschewed emphasizing the effects of
situations, not because situational factors such as power differences or a negotia-
tor's particular audience or constituency are unimportant, but rather because
these situational factors cannot typically be altered by negotiators.
A number of cognitive biases have been reliably uncovered, many of which
are consistent with the more general literature on decision biases (Tversky and
Kahneman, 1974). "A variety of heuristics have been identified which systemati-
cally distort the negotiation process and potentially bias negotiation outcomes.
Four of these cognitive heuristics are framing, anchoring-and-adjustment, avail-
ability, and overconfidence" (Neale and Northcraft, 1990, p. 57). There is a ten-
dency for negotiators to rely too much on available information (Neale, 1984).
"Colorful, dynamic, concrete, and otherwise vivid or distinctive stimuli dispropor-
tionately attract attention and . . . influence decision makers' deliberations"
(Neale and Northcraft, 1990, p. 59). As an example in a negotiating context,
opportunity costs, the cost of an alternative forgone, are much less vivid than out-
of-pocket costs and are therefore not given enough weight in negotiations (North-
craft and Neale, 1986).
The idea of framing comes from prospect theory (Kahneman and Tversky,
1979), which suggests that "decision makers are risk-averse when choosing be-
tween certain gains and the risk of larger or no gains, but are risk-seeking when
choosing between certain losses and the risk of larger or no losses" (Neale and
Northcraft, 1990, p. 57). The same outcome can be framed as a gain or a loss
depending on the individual's initial negotiating position and expectations. Stud-
ies (e.g., Neale and Bazerman, 1985) have shown that positively framed negotia-
tions result in more completed agreements, but negatively framed negotiations
152 New Directions for Organization Theory
negotiators are acting on behalf of a larger constituency. "These studies find that
when negotiators are accountable to constituents who are monitoring their perfor-
mance, they are more likely to engage in noncoopcrative behavior during negotia-
tion" (Kramer, Pommerenke, and Newton, 1993, p. 639). Under conditions of
accountability, there is less likelihood of reaching agreement, concessions arc
reduced, and there is more use of competitive bargaining tactics, Tetlock has
noted, "accountability to constituents . . . induces concern for appearing strong
by refusing to make concessions. People respond by employing competitive bar-
gaining tactics that, while obstacles to resolving conflicts of interest, are quite
effective in protecting their images in the eyes of constituents" (1985a, p. 311).
Kramer, Pommerenke, and Newton (1993) examined the effect of making
social identity and accountability salient in a two-person negotiation context. So-
cial identity was manipulated by having the subjects, MBA students, list all the
ways in which they were similar to other MBAs (high social identification) or to
list all the ways in which they were unique or different (low social identification).
Kramer et al. argued and found that in high social-identity conditions, negotiating
outcomes would be more equal (there would be smaller differences in the scores
earned by the two parties to the negotiation), because "decision making will re-
flect greater concern with the outcomes obtained by the other person" (p. 638).
Accountability also increased the equality of negotiating outcomes, because
"when individuals feel interpersonally accountable to another part)', they are
more likely to use an equality rule when dividing resources" (p. 640).
As Barley has noted, "In daily life, persons are almost always members of
groups whose values and beliefs shape their behavior and cognition. People typi-
cally dispute and bargain as members of families, communities, cliques, and or-
ganizations, not as isolated actors whose judgment is unfettered by social rela-
tionships" (1991, p. 169). Consequently, this line of research is substantively
important for understanding negotiation. It brings elements of the social model
of behavior to the task of understanding this important social process.
In addition to bringing in social context effects, research has explored the
effects of motivational and affective processes on negotiations. Carnevale and Isen
(1986), for example, found that inducing a positive mood facilitated integrative
bargaining. Kramer, Newton, and Pommerenke (1993) reported that individuals
who were higher in self-esteem were more confident and optimistic prior to nego-
tiating and more confident about their doing well in the negotiation process.
Inducing a positive mood also increased negotiator optimism and confidence.
They noted that "if negotiators maintain unrealistically optimistic beliefs about
their abilities and outcomes, they may avoid accepting agreements that they per-
ceive as falling short of their aspirations" (p. 125). Consequently, self-enhancing
bias may be another factor that affects the negotiating process in important ways.
The rise of social cognition as a dominant theoretical perspective in the nego-
tiations realm duplicates its prominence in much of social psychology. But partic-
ularly in the domain of interpersonal influence, the emphasis on cognitive biases
and departures from rational decision making seems too limited. First, in spite of
Neale and Bazernian's (1991) concern with the link between description and
154 New Directions for Organization Theory
prescription, it is not clear that knowing about cognitive biases, or even training
in them and how to overcome them, invariably alters subsequent behavior (but
see Lehman, Lcmpert, and Nisbett, 1988).
In light of the numerous courses and training programs in negotiation, there
is a paucity of research on what determines negotiator effectiveness and whether
or not negotiation training has an effect. Neale and Northcraft (1986) contrasted
the negotiation outcomes produced by professional corporate real estate negotia-
tors with those produced by undergraduate and graduate business students. They
found that there was no difference between amateurs and experienced negotiators
in their susceptibility to the framing bias but that "experts were significantly more
integrative (i.e., were able to achieve agreements of greater joint benefit) than
were amateurs" (Neale and Northcraft, 1990, p. 69). Northcraft and Neale (1987)
examined the anchoring-and-adjustment bias using both experienced real estate
agents and students that asked subjects to estimate the value of a house. The
anchor was the presumed listing price. The study found that experts did not over-
come this decision bias better than amateurs and that, in fact, amateurs were
more aware of the fact that the listing price was influencing their judgments.
Neale and Northcraft concluded, "Experience . . . can lead to superior negotia-
tor performance. However, the truly adaptive superior performance or expertise
comes from having an appropriate strategic conceptualization of the negotiation
process. . . . If cognitive bias in negotiation is preconscious, amelioration of bias
will depend upon the identification of non-cognitive aids to decision making"
(1990, pp. 71-72).
In an experimental study of the effects of assigned goals and training on
negotiator performance, Northcraft, Neale, and Earley (1994) found evidence
that training improved performance and that there was a significant training by
goal setting interaction. In the study, the training manipulation consisted of a
presentation on bargaining and negotiation in which the lecturer "defined negoti-
ation, defined and described advantages of distributive and integrative negotiating
strategics, and developed a model of transforming potentially distributive negotia-
tions into integrative agreements" (p. 263). The data reported in their Table 2
(p. 265) reveals that subjects exposed to training increased their performance sub-
stantially, compared to control subjects who had only the benefit of experience
in a prior negotiation. Thus, there is some evidence that training can have an
effect, although how long and under what conditions remains to be examined.
The effect of experience on negotiation outcomes is more equivocal, which calls
into question how much learning occurs. There would seem to be real value in
seeing what can overcome the cognitive biases that beset negotiators and if over-
coming these biases affects negotiation outcomes.
A second problem with the almost exclusive emphasis on decision making
and social cognition is that the neglect of situational factors because they are not
presumably under the negotiator's control seems an unsatisfactory reason for
avoiding their study. It is far from clear that some relevant situational factors,
including changing the situation to make certain social identities and affiliations
more or less salient, are not capable of being affected. And, the study of processes
of interpersonal influence would benefit from a comprehensive exploration of
Developing and Exercising Power and Influence 155
Conclusion
The topics of power and negotiation again illustrate many of the issues that char-
acterize and influence the development of organization studies. The topic of
power is more likely to be comfortably located in a sociology or political science
department than in a business school, with its emphasis on the economic model
of behavior and on rational decision making. Thus, one might speculate that one
consequence of the changing locus of organization studies has been less emphasis
on power and influence than there might otherwise be, given the importance of
the subject for understanding organizations.
Moreover, the study of power was particularly vigorous in the early 1970s —a
time at which both the strategic contingencies (Hickson et al., 1971) and resource
dependence (Pfeffer and Salancik, 1978) perspectives developed. The 1970s was
a particularly political period, comprising both the end of the Vietnam War and
the associated political protests and the resignation of Richard Nixon as a conse-
quence of the Watergate scandal. The 1980s and 1990s have been much more
politically quiescent; this fact, coupled with a sometimes difficult job market, may
also have made power and influence less likely to be subjects of as much research
attention.
Power as a topic also suffers from the problem of being politically incorrect.
As noted in the discussion of views of control, we prefer a voluntaristic, choice-
based conception of action. Considerations of domination and force, of getting
one's way against opposition—which is, after all, a part of most definitions of
power—perhaps are better left out of sight or discussion. The field's increasing
embeddedness in an economic, rational conception of behavior is nicely coupled
with its renewed emphasis on the individual as contrasted with the situation. Both
rationality and a focus on the individual in isolation more easily permit neglect
of issues of interpersonal influence.
To the extent the field borrows from psychology, then the cognitive emphasis
in psychology becomes evident in the literature on negotiations. Moreover, there
is a practicality to conducting research on negotiations that further pushes the
literature in this direction. Subject pools are virtually nonexistent in business
schools, which means that, if one wants subjects for experiments, one needs to
capitalize on the availability of students in classes, for instance, on negotiation.
Although it is possible to study cognitive biases in class settings, explorations of
the effects of social context are much more difficult to conduct. So, context and
influence of adjacent fields come to affect not only what is studied but how.
7
Organizational Performance
One important goal of organization studies is not only to be able to explain and
predict outcomes within organizations but also to be able to understand why
some organizations fare better than others. There is, of course, substantial mana-
gerial interest in understanding performance, but there are theoretical reasons
as well for attempting to explain why some organizations do better than others.
Performance differences, once generally apprehended, will almost certainly lead
to attempts to mimic the more effective organizations. In that sense, performance
differences create pressures for imitation. Second, to the extent that competitive
markets function, both labor and capital will be drawn to more successful organi-
zations, so that over time a continuing process of natural selection favors higher
performing organizations. Thus, understanding what factors affect performance
can help us predict what the distribution of organizational characteristics will
look like in the future —those characteristics associated with success will be more
frequently represented in the population because of processes of mimetic adapta-
tion and natural selection.
The study of organizational performance is a vast topic and includes contri-
butions from economics as well as organization theory. We focus here on three of
the more prominent organizational approaches for understanding performance-
structural contingency theory and the literature on organizational design, the pop-
ulation ecology of organizations, with its focus on competition and natural selec-
tion as a way of understanding organizational births and deaths, and the recent
work on the effects of management practices, particularly with respect to how
the firm manages its human resources, on organizational performance typically
measured in terms of productivity, quality, or stock market performance. Another
important perspective on performance, emphasizing the organization's network
position and its connection to external sources of support, was considered as part
of the discussion of the social model of behavior in chapter 3 and is therefore not
156
Organizational Performance 157
This approach is, of course, applicable to contexts other than the effects of
board composition (Pfeffer, 1972c; 1973) where it has been demonstrated to have
a statistically significant association with various measures of performance. The
method provides a way of operationalizing the idea of fit, using the actual empiri-
cal results from a sample of organizations under the assumption that, in the aggre-
gate, observed relationships provide a reasonable estimate of the optimal (or at
least better) connection between independent and dependent variables. This
method could be used, for instance, in testing the consonance hypothesis from
structural contingency theory (to be described below), although it has typically
not been. It is a relatively straightforward way of incorporating performance into
studies of other things, and it provides another way of testing the theory. If some
set of variables are presumably related to a given independent variable of theoreti-
cal interest, then it is reasonable to explore whether or not deviations from the
predicted relationships have consequences. To the extent they do, one can have
somewhat more confidence in the underlying theoretical reasoning.
In the mid- to late 1960s and early 1970s, the organizations literature was literally
filled with articles from a structural contingency perspective that maintained: "1.
There is no one best way to organize. 2. Any way of organizing is not equally
effective" (Galbraith, 1973, p. 2). The first statement means that under different
conditions such as variations in strategy (Chandler, 1962), size, technological un-
predictability, or environmental uncertainty, differences in structural arrange-
ments will be observed (e.g., Lawrence and Lorsch, 1967). The second statement,
also known as the consonance hypothesis, means "that those organizations that
have structures that more closely match" or fit "the requirements of the context"
will be "more effective than those that do not" (Pfeffer, 1982, p. 148). Another
implication of structural contingency theory is that "organizations adapt their
structure by moving out of misfit . . . in order to restore effectiveness and perfor-
mance. . . . Much structural change is seen as being positive and productive"
(Donaldson, 1995, p. 33).
At one point, contingency theory seemed to be both widely accepted and
noncontroversial (Schoonhoven, 1981, p. 349). With some notable exceptions
(e.g., Donaldson, 1985), structural contingency theory has since virtually faded
from the research and managerial literature scene. What happened to structural
contingency theory reflects in part characteristics of the field of organization stud-
ies already discussed: (1) an attraction to the new and unique, which makes fol-
lowing through on any cumulative program of research difficult and unlikely—by
the time of Sehoonhoven's (1981) critique the major ideas were more than four-
teen years old and consequently the field's attention was already beginning to
shift; (2) an interest in ideas that are readily translated into action; and (3) a
fascination with economic logic, which is almost invariably noncontingent. As
such, the rise and fall of this brand of organizational analysis provides an example
Organizational Performance 159
of what has happened elsewhere in the field as well as a possible forecast for the
future for other, similar topics.
It is important at the outset to note that not all theories of organizational
structure are contingent. For instance, in economics Williamson (1975) promul-
gated what has come to be known as the M-form hypothesis, which maintained
that the multidivisional (M-form) structure was more efficient than either a hold-
ing company or a functional structure (U-form). Although one might argue that
there was some implicit condition of large size implied —one could scarcely have
an efficient multidivisional structure or perhaps even one at all in very small
organizations—the prediction of the advantages of the M-form were largely non-
contingent: "The organization and operation of the large enterprise along the
lines of the M-form favors goal pursuit and least-cost behavior more nearly associ-
ated with the neoclassical profit maximization hypothesis than does the U-form
[functional] organizational alternative" (p. 150).
The multidivisional form was presumably so effective because it at once sepa-
rated the strategic and capital allocation function of management, performed at
headquarters, from concerns with operational efficiency, located in the divisions.
At the same time, the M-form internalized transactions that would otherwise be
subject to the problems of recontracting under conditions of information asymme-
try and small numbers bargaining and permitted more efficient capital allocation
because of the better information possessed inside organizational boundaries. Wil-
liamson was particularly concerned with problems of commitment to unprofitable
lines of investment in organizations, a problem, he maintained, that would be
overcome by separating the capital allocation process from operational manage-
ment: "existing activities embody sunk costs of both organizational and tangible
types while new projects require initial investments of both kinds. The sunk costs
in programs and facilities of ongoing projects thus insulate existing projects from
displacement by alternatives which, were the current program not already in
place, might otherwise be preferred" (1975, p. 121).
The fact that transfer pricing in multidivisional structures was a highly politi-
cal and contested process filled with its own forms of inefficiency (Eccles, 1985)
seemed to go unnoticed. Armour and 'leece (1978), in a study of firms in the
petroleum industry, did find that adoption of the multidivisional form provided
economic gains until the point at which the form had been so widely diffused
that there was no longer any competitive advantage to be achieved by its adop-
tion. And 'leece (1981), in a study comparing two firms in each of fifteen indus-
tries, one of which had adopted the M-form and one of which had not, also
found a profit advantage for the M-form firms.
There were two problems with the M-form hypothesis as a theory of organiza-
tional structure. First, it did not capture the rnultidimensionality and complexity
of organizational structures. A multidivisional firm still faced a range of choices
about specifically what staff functions, such as human resources, research and
development, and so forth, to place in the divisions and which to centralize
(S. A. Allen, 1978). Furthermore, divisionalization by itself did not capture much
of the range of design variables, including how much discretion to permit the
160 New Directions for Organization Theory
divisions (the degree of centralization) and how to decide what products to put in
one division rather than another (how to coordinate and apportion interdepen-
dence). As Starhuck noted, "organizations are complex entities with myriad mea-
surable characteristics" (1981, p. 181). His review of the so-called Aston studies
of organization structure (e.g., Pugh et al., 1968; 1969) noted that these studies
took more than 1,000 measurements in each organization. Reducing this organi-
zational complexity to a single construct such as divisionalization grossly oversim-
plifies the measurement and meaning of organizational structure.
Second, the theory did not fit the facts as they were being uncovered by
empirical studies of organizational structures — namely, that structure seemed to
vary quite systematically with an organization's strategy, size, technology, and con-
ditions of the organization's environment (e.g., Pugh et al., 1969; Child, 1973;
Child and Mansfield, 1972). Common sense suggested that the appropriate orga-
nizational arrangements must surely depend on what is being organized and the
environment in which that organization has to operate; consequently, the ideas of
structural contingency theory assumed prominence.
Figure 7-1 presents an overview of the major variants of structural contin-
gency theory. The facets of organizational structure that have been examined
include the size of the administrative component, formalization, centralization,
the degree of both vertical and horizontal differentiation, and the degree of task
specialization and complexity. The factors associated with variations in organiza-
tional structure have included strategy, technology, particularly the technology of
production, organizational size, and dimensions of the organization's environ-
ment, including the degree of competition and the amount of uncertainty or
change in the environment.
There has been more support for the proposition that structure varies predict-
ably across organizational contexts than there has been for the consonance hy-
pothesis, which maintains that a match between context and the structure best
suited to that context is associated with enhanced organizational performance.
Considering first some of the evidence on the effect of context on structure, there
is evidence that organizational size has been among the most important predictors
of structural variation. Blau (1970) argued that size produced structural differenti-
ation but at a diminishing rate as size increased. Meyer (1972b), in a longitudinal
study of state and municipal finance departments, found that siz.c did cause struc-
tural differentiation. Other general findings from the literature on the effects of
si/e are: larger organizations tend to be more formalized (Meyer, 1972a; Hall,
Haas, and Johnson, 1967), more specialized (Child, 1973; Pugh et al., 1969), and
less centralized (Meyer, 1972, Blau, 1973). Child (1975) and Khandwalla (1973)
found that larger size was associated with increasing levels of the degree of bu-
reaucratization.
Scott (1992) has reviewed the literature on the effects of technology on struc-
ture, where technology has most frequently been conceptualized in terms of its
routincness, variability, or unpredictability. The basic argument is that nonrou-
tine, variable, or unpredictable technologies require more flexible organizational
arrangements such as decentralization and less formalization (Perrow, 1970b;
Hage and Aiken, 1969). The literature on the environment has also emphasized
uncertainty or variability (e.g., Duncan, 1972). Burns and Stalker (1961), in what
is probably one of the first studies in the structural contingency tradition, found
that a more mechanistic or bureaucratic structure was appropriate for more stable
and certain environments while a more flexible, organic structure was both ob-
served more frequently and was more successful in more dynamic and uncertain
environments. There is also some evidence that competition or the degree of
stress emanating from the environment is associated with more centralization of
control (Pfefferand Leblebici, 1973; Khandwalla, 1973).
There have also been a number of demonstrations of the relationship be-
tween strategy and structure. Rurnelt (1974) found that more diversified firms
were more likely to employ a divisionalized structure. Undiversified firms were
more likely to retain a functional structure. Moreover, the evidence is that firms
diversified first and then changed their structure, not the other way around, as
would be expected if structure caused strategy (e.g., Rumclt, 1974; Channon,
1973) rather than the reverse.
Schoonhoven (1981) has detailed the extensive theoretical and empirical
problems that beset the test of the consonance hypothesis. This hypothesis has
received support in some studies (e.g., Woodward, 1965) but not in others (Mohr,
1971; Pennings, 1975). Schoonhoven noted that (1) "contingency theory is not a
theory at all ... it is more an orienting strategy or metatheory" (p. 350); (2)
there is a lack of clarity by contingency theorists that they are, in fact, predicting
interactions; (3) "theoretical statements fail to provide any clues about the specific
form of the interaction intended" (p. 351); (4) relationships are assumed to be
linear, even though there are many reasons to suspect nonlinearities; and (5)
there is an assumption that effects are symmetrical, even though, again, there is
reason to suspect asymmetry in the interactions.
162 New Directions for Organization Theory
1. The generalizations should inform the users not only what is likely to
happen under the specified conditions but how to create the condi-
tions and actions in the first place. . . . (p. 5)
2. Propositions that are intended to be used . . . if used correctly should
lead to the predicted consequences, and not to others that are counter
to those predicted. . . . (p. 8)
Organizational Performance 163
Organizational Ecology
as return on average assets. This was a period in which the regulatory rules were
evolving, both nationally and also in Illinois, which was modifying its limitations
on branch banking to permit more branching. Barnett et al. examined the relative
performance of rmiltiunit and single-branch banks. They found a paradox: unit
banks suffered from competition, but the survivors were more successful competi-
tors from learning how to survive under intense competitive pressure. By contrast,
"branch systems . . . circumvent and avoid competition by attaining positional
advantage — but then forego [sic] the evolutionary benefits competition brings"
(p. 25). The authors argued that "rather than strategy and structure driving com-
petitiveness, it is competition that drives evolution, which is then shaped by the
strategies and structures of organizations. Ironically, the more they mollify selec-
tion pressures, the more that strategics and structure allow organizations to survive
regardless of their ability to learn from the market" (p. 24).
Ecology empirically defines organizations that are independent, competitive,
or mutualistic with each other in terms of the effects of population density on
births and deaths: "By ecological conception, shared fates among organizations
indicate interdependence. When organizations negatively affect one another, they
are competitive. When they enhance each other's viability, organizations are mu-
tualistic" (Barnett and Carroll, 1987, p. 400). Thus, an ecological view of techno-
logical interdependence examines the effects of density of organizations with a
given technology on the failure rates of organizational populations —if the pres-
ence of more organizations with a given technology increases failure rates, then
there is evidence of competition; if they decrease failure rates, there is evidence
of mutualism (Barnett, 1990). The nature of the interdependence among organi-
zations is, therefore, defined by its consequences.
Hannan and h'reeman (1984) argued that inertia itself was produced by envi-
ronmental selection pressures and was not simply the consequence of managerial
incompetence. They noted that "in modern societies organizational forms that
have high levels of reliability and accountability are favored by selection pro-
cesses. Reliability and accountability of organizational forms require that the orga-
nizational structure be highly reproducible" (Singh and l.umsden, 1990, p. 168),
which favors structures that don't change.
The basic ecological argument is that organizational forms that are compara-
tively more suited to the environment or niche within which they are operating
will fare better —that is, they will tend to exhibit a higher founding rate and a
lower mortality rate. It is important to note that the operative phrase is "compara-
tively" more fit, for unlike economics, ecology makes no claims about optimality
or even progress. If a better form is not present in the population, for whatever
reason, it cannot triumph in the selection process, which means that there is no
guarantee that the organizational forms that scern to be winning are optimal.
Moreover, because survival depends on the fit or match between the characteris-
tics of organizations and their environments, which constantly change, what is a
better or worse organizational form obviously changes also. Thus, in the popula-
tion ecology of organizations, there is never a best way to organize — only a way
that has outperformed rivals in the past, given the conditions of the environment
Organizational Performance 165
that prevailed at that time. Finally, it is important to note that while organiza-
tional ecology deals with selection pressures, those selection regimes may have
political elements. There is no assurance that selection is only on the basis of
efficiency or economic performance (Hannan and Freeman, 1989).
The empirical results of the numerous organizational ecology studies are
readily summarized. First, there is a liability in newness (Stinchcombc, 1965), in
that new organizations tend to fail at a higher rate than older ones. Singh and
Lumsden summarized the arguments about why new organizations have higher
failure rates:
Young organizations and the individuals in them have to learn new roles. . . .
A significant amount of time and effort has to be expended to coordinate these
new roles for the individual actors and in their mutual socialization. . . . In
dealing with external clients, customers, and other relevant actors, new organiza-
tions are forced to compete with existing organizations that have well-established
client groups who are familiar with the organization. (1990, p. 168)
This is an interesting and nontrivial prediction, when one considers the fact that
older organizations arc likely to have greater inertia and be more difficult to
change. A prediction founded just on considerations of adaptability, then, would
predict a higher failure rate for older organizations, particularly in more rapidly
changing environments that would require more change.
This age dependence of organizational mortality holds even when size is
controlled —it is not the case that age dependence occurs simply because new
organizations also tend to be smaller and smaller organizations disappear more
frequently, although there is a liability of smallness (Aldrich and Auster, 1986).
Smaller organizations have more difficulty raising capital, face diseconomies of
scale in dealing with government regulations, and face problems in attracting
labor in competition with larger organizations, because they cannot offer the ca-
reer prospects and stability of larger organizations. However, in a study of health
maintenance organizations, Wholey, Christiansen, and Sanchez (1992) found
that the liability of small size depended importantly on organizational form and
was not necessarily a universal effect.
Carroll (1983) examined the liability of newness in a diverse sample of firms
in fifty-two quite different industries and found support for the idea. Carroll and
Delacroix (1982), studying Argentine and Irish newspapers, found that failure
rates were related to age, and Freeman and Hannan's (1983) study of restaurants
also observed age-dependent mortality. Freeman, Carroll, and Hannan (1983) ex-
plored the liability of newness using data on unions and semiconductor firms, an
analysis continued in Hannan and Freeman (1989). The argument that many
organizational routines and practices entail tacit knowledge and skill and that,
until that knowledge is developed and those interdependencies effectively coordi-
nated, organizations are at greater risk for failure implies that reorganizations or
other substantial changes (such as a change in the leadership of the organization)
should reintroduce at least some of the initial mortality risk, until the organization
learns to operate under the new conditions. And, indeed, there is evidence that
166 New Directions for Organization Theory
third, related problem arises beeause of the concern in much of the literature
solely with demonstrating effects of density rather than fully understanding all the
forces affecting organizational birth and death. This leads to models that may
underspecify causes of birth (or failure) that vary with density and, as a conse-
quence, lead to estimated density effects when there are, in fact, no such effects
in better specified models. Wholey, Christiansen, and Sanchez (1993) examined
the founding of health maintenance organizations. They reported that:
There are no density effects in the fully specified model, which is surprising
given the consistent density effects reported by population Geologists. . . . We
found nonmonotonic density effects that are consistent with ecological research
in much simpler models. . . . This suggests that the nonmonotonic density ef-
fects in simpler models are a consequence of specification bias with density serv-
ing as a proxy for other processes affecting entry. Density effects disappear when
we include comprehensive measures for these processes, (p. 178)
As Baum (1995, p. 178) has noted, density dependence ideas treat competi-
tion as a property of the population and consequently assume that each member
of the population contributes equally to the competitive environment and experi-
ences competition similarly. This seems unrealistic on its face, as there are organi-
zations of different sizes that overlap in their market niches with competitors to
varying degrees, and consequently the competitive effects should not necessarily
be the same. This insight has led to refinements or modifications in the basic
density dependence ideas. Baum and Mezias (1992) found that the failure rate of
Manhattan hotels was greater the greater the intensity of competition from hotels
of similar price, location, and size. Carroll and Wade (1991) found that an in-
crease in density measured at the local geographic level had a greater effect on
the failure rate. Baum and Singh (1994) developed the concept of overlap density
and used it to study the failure rate of day care centers, measuring population
overlap density in this instance in terms of the ages of children each day care
center was licensed to serve.
Barnett and Amburgcy (1990) argued that what might be important was not
just population density but also population mass—the aggregate of the sizes of
organizations in a population. They noted that larger organizations might gener-
ate more intense competition because such organizations enjoyed economies of
scale and would deplete common resource pools to a greater extent. On the
other hand, larger organizations may be less flexible and adaptive. They defined
population mass as "population density with each organization weighted by its
size" (p. 83). Their study of telephone companies in Pennsylvania found a mutu-
alistic rather than a competitive mass effect, but the study measured mass "by the
total number of subscribers of all existing companies in a given year" (p. 90).
This measure of mass-dependent competition was not consistent with their own
definition and could be argued to be a measure of total market size, not the size
distribution of the competitors. Baum (1995, p. 181) has summarized the research
on mass-dependent competition by noting that the empirical results were mixed.
While density- dependence ideas presume that each organization in a population
competes with all other organizations, in models of size-localized competition,
168 New Directions for Organization Theory
the basic argument is that the failure rate of organizations increases with the
number of organizations of similar size in the population (p. 182). This prediction
comes from the assumption thai organizations of different sizes compete primarily
with similarly sized organizations that have similar strategies and cost structures.
Closely related to size-localized competition is the idea of resource parti-
tioning. In a study of newspapers, Carroll (1985) argued that competition among
large, generalist organizations increased their failure rate but lowered the failure
rate of small, specialist organizations that could serve small market niches ne-
glected by the large, generalist firms. Carroll did observe that as the newspaper
industry become increasingly concentrated, the failure rate of large, generalist
newspapers increased while the failure rate of small, specialist newspapers (often
published in foreign languages) actually fell. Similarly, Carroll and Swaminathan
(1992) observed that as the level of concentration increased in the American
brewing industry, the failure rate of microbreweries decreased. Baum (1995),
studying the New York City hotel industry, found empirical support for each of
these bases of competition and also noted that which form of competition pre-
dominated depended on period—there were temporal differences in the stages of
competition. Thus, the empirical evidence suggests that competitive dynamics are
better modeled using more refined measures of organizational similarity (size,
price, geographical location, target population served, and so forth) rather than
just aggregate population density.
What we have learned from the various ecological studies is that there is a
liability in being new or young, a liability that recurs to some extent when there
are reorganizations or important successions. There is also a liability to being
small. We have learned that foundings are more likely and organizational mortal-
ity is diminished the fewer similar organizations there are and, conversely, that
foundings are less likely and mortality is greater the more dense the population
becomes. We have also learned in numerous studies that both foundings and
mortality have a curvilinear relationship with population density because of the
joint effects of legitimation and competition. The studies of particular industries
have also examined more specific effects such as political and economic condi-
tions that are relevant for understanding organizational birth and death in the
particular population.
As a theory of performance, population ecology focuses attention on factors
that may be substantively important but over which there is virtually no organiza-
tional control (e.g., age, size, and the density of competing organizations) except
to the extent organizations can choose their competitive niche. But once that
niche is chosen, given presumptions of inertia and the very real constraints that
make entering and leaving market sectors difficult, a given organization's fate is
largely in the hands of others and selection forces that are outside of its control.
And even how to choose a niche is not clear from the empirical results. In the
study most directly focused on performance (Barnett et al, 1994) and strategy,
there was a paradoxical result: the very strategies that insulated organizations from
competitive pressure and therefore enhanced their life chances also insulated or-
ganizations from the selection pressures that foster learning. The choice posed is
Organizational Performance 169
600 percent difference between the most effective lines, which were shut down 2
percent of the time, and the least effective, which were down about 13 percent
of the time.
There is also evidence that an important portion of the variation in economic
performance is associated with differences in human resource practices. MacDuf-
fie (1995), using data from the MIT study of the world automobile industry,
observed a strong positive relationship between his measure of production organi-
zation and both quality and productivity. His production organization index in-
cluded scales measuring the use of buffer inventories, work systems (including
the use of work teams and suggestion systems), job rotation and decentralization,
and human resource management policies (including contingent compensation,
an emphasis on training, and limited differentiation in status). Previous but less
comprehensive studies in the automobile industry had already demonstrated im-
portant effects of, for instance, participation in suggestion programs, involvement
in quality of work lite programs, and worker attitudes on both productivity and
quality (Katz, Kochan, and Gobeille, 1983; Kate, Kochan, and Weber, 1985). The
magnitude of some of the bivariate effects are astonishing. Katz, Kochan, and
Weber (1985, p. 519) reported a bivariate correlation of 0.73 between participa-
tion in suggestion programs and product quality in their study of twenty-five Gen-
eral Motors' automobile assembly plants.
Kravetz (1988) examined the connection between human resource policies
and financial performance for 150 of the 500 largest companies listed in Forbes.
His measure of human resource progressiveness included the extensiveness of
career development and training, participative management, flexible work sched-
ules, and the degree of emphasis on people in the company culture. He found
that the more progressive companies enjoyed 64 percent greater sales growth, 61
percent higher profit margins, four times the rate of increase in profits, and a
larger increase in earnings per share than companies using less progressive human
resource policies. Reviewing 131 field studies in North America between 1961
and 1991 that assessed the effects of changes in work systems, Macy and Izumi
concluded that "there seems to be large overall performance improvements ob-
tained from work innovation . . . and organizational development efforts" (1993,
p. 265). A metaanalysis of forty-three studies of various forms of worker participa-
tion— participation in decision making, mandated co-determination, profit shar-
ing, worker ownership (either through employee stock ownership or individual
worker ownership of the firm's assets), and collective ownership —concluded that
except for co-determination, all of the forms of worker participation were posi-
tively associated with productivity (Doucouliagos, 1995). Moreover, the effects of
participation were larger in firms owned and controlled by workers compared to
firms in which participation was granted by owners or managers to workers.
Ichniowski, Shaw, and Prennushi's (1993), study of thirty integrated steel pro-
duction plants, using the percentage of time the line actually operated as the
dependent variable, found that, controlling for technological factors, human re-
source systems had a significant effect on performance. Studying thirty steel min-
imills, Arthur (1994) observed a statistically and substantively significant relation-
ship between how an organization managed its work force and measures of both
Organizational Performance 171
mean, only 60 percent survived. The effect of rewards was even larger. The sur-
vival rates for firms one standard deviation above and one standard deviation
below the mean were 87 percent and 45 percent, respectively. Welbourne and
Andrews's results demonstrate yet again that human resource practices are both
statistically significantly related to important performance outcomes and that
these effects are substantively important.
Husclid's results indicated that there was little evidence for complementarity
either among the various internal practices or between the internal practices and
the firm's strategy. The data showed that the effects of the various high-
performance practices were largely additive and noncontingcnt on a particular
strategy. Ichniowski et al. (1993) did find evidence of complementarity among
the various practices, and Arthur (1992) had found a relationship between strategy
and the use of high-commitment or cost-reduction work practices in his study of
steel minimills. Complementarity and the effectiveness of practices regardless of
strategy may be contingent on the particular firms and their environment.
Although the various studies differ somewhat in what they include when they
measure high-performance or high-commitment work practices, there is substan-
tial overlap in the commonly accepted meanings of these terms. Figure 7-2 pre-
sents dimensions of work practices from a number of studies and indicates that
there are a number of common themes across the studies. Specifically, high-
performance work practices are characterized by higher levels of wages, employee
skills, effort in selecting employees, and training and a set of practices associated
with the devolution of power, including the use of quality circles or self-managed
teams (Katzenbaeh and Smith, 1993; Mohrrnan, Cohen, and Mohrrnan, 1995;
Seaman, 1995), fewer job classifications, more information sharing, and fewer
supervisors. Having better-trained and more-carefully screened employees working
a system that permits and, indeed, encourages them to use their skills results in
higher levels of organizational performance.
mance' work organizations" (Karr, 1990, p. A4). Diffusion is equally slow in the
United Kingdom. R. Loeke reported that "no more than 2% of all establishments
with more than 25 workers have quality circles or problem-solving groups. Team
work and major alterations in job content are even more rare" (1995, p. 16).
In the textile industry, the existing bundle system of production, which is
characterized by low wages, a high rate of occupational injury, and little career
mobility, is favored by almost no one. Dimlop and Weil (1992) noted that many
industry observers favored an alternative system of production called modular pro-
duction, which could potentially solve problems of absenteeism, turnover, and
labor shortages. Nonetheless, the diffusion of the modular production system and
the abandonment of the bundle system lias been slow:
In 1985, 97% of all manufacturers used the bundle system, according to industry
surveys. Although this figure decreased to 90% by 1988 and to 82% by 1992,
given the economic pressures and the fact that industry publications were filled
with articles advocating the new system, one might have expected more change.
The components of the bundle system, such as individual (rather than group)
piece rates and narrow (as opposed to broad) training also maintained their domi-
nance in the industry. (Pfeffer, 1994, p. 83)
additions or expansion. They also found no effect for previous layoffs on change
in the use of these ways of organi/ing work. They argued that the absence of an
effect might reflect two counteracting forces —the tendency for layoffs to, on the
one hand, create distrust and fear, which would make change difficult, but on
the other hand to create the perception of the need for change and thereby un-
freeze strongly held attitudes, thereby making change easier.
The Pil and MacDuffie study did not directly examine many of the factors
that have been discussed as affecting the diffusion of high-performance work prac-
tices—for instance, managerial ideology, the environmental context, which obvi-
ously varies dramatically across the various nations in which their plants were
located, or even the tendency for high-performance practices to diffuse more
within a single corporation or geographic area compared to across organizational
or geographic boundaries. The literature indicates that there arc a number of
potential explanations for why some firms adopt high-commitment work practices
while others do not. What is clear is that this is an increasingly important area
for research.
Conclusion
Much, although not all, of the literature on organizations seeks to address mana-
gerial problems — how to enhance organizational performance, how to exercise
more effective control over behavior in the work place, how to create and manage
organizational cultures, how to identify and develop leaders, and so forth. The
dominant approach in this literature has been functionalist:
The functionalist paradigm has provided the foundation for most modern theory
and research on the subject of organizations. . . . The perspective . . . encour-
age^ us to see the role of values as a separate variable in the research process.
. . . Functionalist theory has typically viewed organization as a problematic
phenomenon, and has seen the problem of organization as synonymous with
the problem of'efficiency' and, more recently, of'effectiveness'. (Morgan, 1990,
P. 15)
Most of the undesirable features of modern capitalistic society are not mentioned
by organization theorists, presenting the appearance of value neutrality, but in
actuality masking a politically conservative bias. This compromises the social
function of the field to the point where organizational theory serves primarily the
dominant interest of capital, rather than society at large. (Jermier, 1982, p. 204)
we arc intellectual captives of the organizations we study, and thus are burdened
with some unproductive assumptions. . . . We have failed to pay our dues to
society by devoting . . . roughly a quarter of our efforts to applying our organisa-
tional expertise to pressing public problems and in particular, those of the power-
less. . . . I question the assumption that efficiency can be addressed by exam-
ining survival, legitimacy, growth or profits, thus neglecting the multiple
stakeholders within and without the organisation with quite different notions of
. . . goal attainment. . . . We should question our evolutionary bias that sees
big organisations as more complex, efficient, and selected by an indifferent envi-
ronment, and instead sec organisational birth, death and change as a process of
decimation of forms and reduction of adaptability. (1992, p. 371)
The Marxist critique of organization theory was pursued most vigorously in the
1970s. In some sense this is not surprising, as the 1970s saw the end of the Viet-
nam War, a very politicizing event for American culture, Watergate and the resig-
nation of President Nixon, and the economic stresses associated with oil shocks,
inflation, and recession. It was a time of heightened political awareness and a
questioning of the institutions of authority. Marxist theory provided a useful theo-
retical lens for that critique:
There are two topic domains in which Marxist approaches pose the most
direct challenge to traditional organization theory—in the structuring of work
within enterprises and in the understanding of relations among enterprises and
between enterprises and the state and the reasons for those relations. We consider
each of these topics in turn.
control, not efficiency, is the objective of organizing arrangements and that when
there are trade-offs involved, efficiency concerns are frequently subservient to the
achievement of control over the labor process. Marglin's (1974) analysis of the
development of hierarchy during the industrial revolution, Stone's (1973) descrip-
tion of the development of job ladders in the steel industry m the late nineteenth
century, and Clawson's (1980) review of the demise of the inside contracting and
putting out system and their replacement by hierarchically structured employ-
ment relations all argue that changes in the structure of the employment relation
emerged "as much from the desire to effect organizational control as from the
need to apply continually advancing technology to production" (Goldman and
Van Iloutcn, 1977, p. 109). Edwards has argued that several factors enter the
evaluation of profitability, including "the extent to which any technology provides
managers with leverage in transforming purchased labor power into labor actually
done" (1979, p. 112). Second, a critical or Marxist perspective argues that many
of the prescribed (and often acknowledged as effective) mechanisms for achieving
control of the work process are not benign and achieve their results at some
substantial cost to the individuals working in organizations.
Empirically testing whether control is instituted to enhance efficiency or for
other reasons is obviously difficult. One can observe outcomes or behavior much
more unambiguously than motives. Four lines of evidence and argument may be
relevant to the debate. First, there is evidence, already discussed, that there is less
diffusion of high-commitment work practices, and particularly practices that in-
volve fundamental changes in the power structure and in control of work, than
one might expect given the effects of such practices on productivity and perfor-
mance (e.g., Levine and Tyson, 1990; Pfeffer, 1994). Data from a General Ac-
counting Office survey of the Fortune 1000 reported by Lawlcr et al. (1992)
showed that self-managed teams and gain sharing (sharing financial benefits from
increased performance accompanied customarily by some degree of control over
the work process) were work-place changes that had diffused the least. By contrast,
survey feedback and all salaried pay had diffused the most completely. Note that
changing pay from an hourly to a salaried basis and collecting survey data and
feeding it back involve less change in the control of work than going to self-
managing teams. The data also indicate that the differences in diffusion cannot
be explained by differences in perceived effectiveness of the changes. Similarly,
the failure of decentralization to readily diffuse (Levine and Tyson, 1990) is con-
sistent with a desire to maintain control.
Second, there is evidence that employers pursue control over the labor pro-
cess and contest that control even when there arc few or no apparent efficiency
reasons for doing so. For instance, although Freeman and Medoff (1984) have
documented the frequently positive and seldom harmful effects of unionization
on productivity and profitability (see also Pfeffer, 1994, chap. 7), employers in
the United States have resisted unionization vigorously (Kochan, Katz, and
McKersie, 1986).
L. J. Griffin, Wallace, and Rubin (1986), studying employer resistance to
unionization in the early 1900s, argued that such resistance was based impor-
tantly on a desire to achieve control over the labor process. They wrote, "Union-
Organizations from a Critical Theory Perspective 181
diffusion of these practices means that skill and educational requirements have
gone up where they have been adopted and have either stayed the same or gone
done where more Taylorist forms of work organization persist. The mixed empiri-
cal results suggest that a more finely variegated theory of occupational change is
required to understand this aspect of the evolution of work in America.
The inconsistent empirical results also highlight the problems in the defini-
tion and measurement of skill levels. Spenner has argued that there are at least
two dimensions to skill: substantive complexity ("the level, scope, and integration
of mental, manipulative, and interpersonal tasks in a job" [1990, p. 402]) and
autonomy-control ("the discretion or leeway available in a job to control the con-
tent, manner, and speed with which tasks arc done" [p. 403]). As he noted, "there
is far more consistent evidence of dcskilling with respect to the effect of techno-
logical change on levels of autonomy-control than on substantive complexity"
(p. 404). Barley's (1988) review of this issue also highlights the important distinction
between autonomy and skill. Thus, the practice of health maintenance organiza-
tions and insurance companies to require preauthorization for hospitalization and
some expensive medical tests and procedures clearly limits physicians' autonomy,
but it docs not change the skill or knowledge required to do the work of a doctor.
Bailey noted that the "direct measurement of skill has proved to be extremely
difficult" (1995, p. 28) and that even when such observations are made, the re-
sults measure the skills available and used rather than those that are optimal for
doing the work. lie noted that two indirect approaches to measuring skill levels
were (1) measuring the relative earnings of workers with different levels of educa-
tion and (2) measuring the growth of different occupations with presumably dif-
ferent skill requirements. The change in relative wages during the 1980s was
consistent with the view that skill requirements were increasing, as the earnings
difference between college graduates and high school graduates increased from
13 percent in 1979 to 38 percent by 1987 (p. 29). The analysis of the entire
occupational distribution (as contrasted only with the ten fastest growing jobs)
also suggests that "the new jobs that will be added over the next decade are
disproportionately those that currently use more highly educated incumbents"
(p. 30). Educational requirements may represent credentialism rather than actual
learning required to do a task (Collins, 1979), so analyses relying on these indirect
measures are inevitably somewhat problematic.
In his review of the literature on technical change, automation, and skill
levels, Barley concluded: "current theories of technology and work are either too
brutish or too brittle to capture the subtle but multiple ramifications of technical
change. . . . Between the stark degradation heralded by the neo-rnarxists and the
bright industrial Utopia envisioned by the sociologists of automation, must lie a
more nuanced depiction of events" (1988, p. 72).
Some of the earliest Marxist works on the labor process (e.g., Blauner, 1964)
noted the alienating effects of achieving control over the labor process — which
almost inevitably entails separating the planning of work from doing the work
and a process of deskilling (Braverman, 1974). As competitive exigencies have
required the involvement and commitment of employees' minds as well as their
bodies, there has been a trend toward using high-commitment work practices
Organizations from a Critical Theory Perspective 183
(Osterman, 1994). These practices often entail creating self-managing teams and
increasing involvement and participation in the work place.
Yet these changes are not necessarily favorable in their effects on employees.
As Barker (1993) demonstrated in an ethnographic study of an organi/ation mov-
ing from hierarchical to team-based control, the new, value-based and peer-
monitored control system was more powerful and complete in its effects than the
hierarchical system it replaced. Adler's (1992) study of New United Motors pro-
vided ample evidence that the control of work under the new system was much
more complete and efficient than under the old. The workers were possibly less
alienated and more motivated, however, because now they were supervising each
other. As M. Parker and Slaughter wrote in describing the NUMMI system and
its reduction of absenteeism, "all the difficulties of one person's absence fall on
those in daily contact with the absentee— the co-workers and immediate supervi-
sor—producing enormous peer pressure against absenteeism" (1988, p. 43). Par-
ker and Slaughter referred to the system as management by stress. The question
is whether the team model is "creating an era of cooperation between workers
and management that is rooted in coercion or consent" (Graham, 1995, p. 1).
Graham's (1995) participant observation study of a Subaru-Isuzu plant in the
United States documented the tremendous pace of work and how that pace puts
enormous pressure on employees, creating stress and even physical injury. Her
study also documented how production exigencies frequently overrode formal
company policies of egalitarianism and how, even in a nonunion environment,
the work place demands provoked resistance. Garrahan and Stewart, interviewing
a small number of workers from Nissan's plant in the United Kingdom, con-
cluded that "the Nissan Way rests upon control through quality, exploitation via
flexibility, and surveillance via tcamworking" (1992, p. 139). A study of a small
number of employees at the Mazda plant at Flat Rock, Michigan (Fucini and
Fucini, 1990), also found dissatisfaction with the production system. A study pre-
dicting individual employees' attitudes toward lean production (Shadur, Rodwell,
and Bamber, 1995) found that commitment to the company was the single best
predictor, with attitudes toward the speed of the line being second in importance.
There is irony in the fact that the Taylorist system, with its emphasis on the
analysis of work processes, has been perfected not by separating the planning of
the work from its doing, as Taylor had proposed, but rather by having the workers
participate in their own control and design their own work processes. Parker and
Slaughter found that in the NUMMI system, every move a worker made was
specified in greater detail than ever before. Because supervisors know the produc-
tion process well from working on the line themselves, and because workers are
involved in a process of continually improving (increasing the productivity) of
their own jobs, there is a much more complete form of control achieved than in
a more traditional hierarchy.
The Marxist analysis of the labor process soon came under criticism even
from critical theorists (e.g., Burawoy, 1985; M. Reed, 1990). First, some argued
that the representation of managers as all having the same interests and acting in
a unified fashion was incorrect. Second, many writers noted that, whatever the
interests and intent of managers and capitalists, workers were not as powerless as
184 New Directions for Organization Theory
dination. Palmer concluded that rather than director interlocks linking organiza-
tions, organizations constituted arenas that linked the elite together. Palmer,
Friedland, and Singh (1986) found that neither competitive constraint nor inter-
industry transaction interdependence accounted for the reconstitution of broken
ties, providing further evidence that director interlocks did not apparently serve
interfirm coordination.
Another domain relevant to class solidarity is whether or not the business
elite took coordinated political activity in spite of the apparent divergence of orga-
nizational interests around specific issues. For instance, Whitt (1979; 1980) stud-
ied five transportation issues determined by referenda in California during the
period 1962-1974 and noted, "Even though there is good reason to expect that,
for each issue, some companies would favor it and some would oppose it, the
money in every case is virtually all on one side or the other of the issue" (1980,
pp. 105-106). Whitt also observed a strong correlation between measures of firm
size or stake in the issue and the amount contributed, which he took to indicate
coordination of action. Dunn's (1980) study of the Weyerhaeuser family office
provided an example of one possible coordinating mechanism and how it worked,
while Dornhoff (1974) has argued for the importance of social ties realized in
clubs and other such settings for achieving intraclass coordination. There is evi-
dence that corporate interlocks also help to structure business elite activity and
that unit)' in perspective extends to political candidates as well as to the referenda
studied by Whitt (e.g., Clawson and Neustadtl, 1989; Clawsoii, Neustadtl, and
Bcarden, 1986; Mizruchi and Koenig, 1986).
The final form of evidence mustered to show the importance of class has
been demonstrating a structure to the ruling class (Domhoff, 1967; 1970; 1974).
The argument is that there is an inner circle of elite business leaders that hold
multiple directorships, articulate classwide business interests, and make decisions,
for instance, on political contributions, with these general (as opposed to particu-
lar corporate interests) in mind. This structure is manifested in the frequency and
centrality of both directorship ties and ties to other important organizations such
as government, prestigious foundations and charities, and policy influencing bod-
ies such as commissions and associations (Useem, 1979). Useem (1984) studied
the inner circle in both the United States and the United Kingdom, examining
board of director interlocks, philanthropic and political contributions, service on
nonprofit boards, social origins, and membership on important government advi-
sory committees. He wrote:
Central members of the inner circle arc both top officers of large firms and
directors of several other large corporations operating in diverse environments,
'['hough defined by their corporate positions, the members of the inner circle
constitute a distinct, semi-autonomous network, one that transcends company,
regional, sectoral, and other politically divisive fault lines within the corporate
community. The inner circle is at the forefront of business outreach to govern-
ment, nonprofit organizations, and the public, (p. 3)
The "inner circle" distinction is critical for this line of argument, because
there is too much heterogeneity otherwise in business behavior to argue convinc-
ingly for the importance of class-based interests and action. But the inner circle
]86 New Directions for Organization Theory
argument makes logical sense. In any social structure, including individual organi-
zations, those admitted to the highest levels of power and prestige are individuals
who have been carefully filtered (J. C. March and March, 1977) and who exem-
plify and uphold the organization's central values and interests. There is no rea-
son to believe that obtaining high-level positions in a structure that transcends
organizational boundaries would not similarly depend on one's ability and willing-
ness to advocate positions that serve the network as a whole.
Yet the idea of the unity of the ruling class is challenged by the plethora of
hostile takeovers (Hirsch, 1986) as well as by the greater activism on the part of
boards of directors in the 1990s to dismiss poorly performing CEOs—friends don't
raid each other or fire each other. It is possible that both positions are partly
correct. There may be an inner circle and an attempt to develop collective inter-
ests that transcend the parochial concerns of individual organizations (Useem,
1984, p. 5). At the same time, there are obviously pressures, including those from
institutional investors (G. F. Davis and Thompson, 1994), that challenge this cozy
relationship and threaten its stability. Which tendency—-classwicle coordination
or the pursuit of individual organizational interests —dominates at any point in
time is contingent on a number of environmental factors, and as Davis and
Thompson (1994) noted, it is not just the ruling elite that can organize for collec-
tive action but their critics and investors as well.
As an example, Pfeffer and Davis-Blake (1987) showed that salaries for both
male and female college administrators decreased as the proportion of women in
administrative positions in a given college or university increased and that in
colleges and universities that paid less, the proportion of women administrators
increased over time more than in settings in which pay was higher. The explana-
tion was that women's work was socially devalued, and it had become taken for
granted or institutionalized to pay less for "women's jobs." Galas and Smircich
critiqued the study not for its findings but for its apparent silence about implica-
tions. They wrote: "they don't use institutional theory to question this situation —
which defies the expected positive social consequences of an increased number
of females in professional activities" (1992, p. 231).
Critical theory maintains that social science is never neutral, particularly with
respect to issues of power, and that this needs to be more explicitly acknowledged
in both theoretical development and empirical work. Thus, critical theorists argue
that even texts that do not mention gender—or, perhaps, particularly works that
do not seem to directly implicate gender—by this very omission contribute to the
domination of women (e.g., Mumby and Putnam, 1992; Martin, 1994).
Feminist literature has criticized the separation of the public sphere of work
and the private sphere of home and family life (Olsen, 1983; Martin, 1994).
Although often portrayed as being gender neutral, some argued that "the reifica-
tion of the public-private dichotomy gives organizations a justification for refusing
to deal with the ways that home life and work life are inextricably intertwined"
(Martin, 1994, p. 409). Because the division of home and child-rearing respon-
sibilities are not gender neutral, there are differential effects of being married and
having children on women's and men's careers (e.g., Pfeffer and Ross, 1982).
Writers have argued that the very language describing organizations has gen-
der connotations. Acker noted that "the organization itself is often defined
through metaphors of masculinity of a certain sort. Today, organizations are lean,
mean, aggressive, goal oriented, efficient, and competitive but rarely empathetic,
supportive, kind, and caring" (1992, p. 253). Similar arguments have been made
about the literature on leadership (Galas, 1993); these arguments note that the
language used to describe effective leadership behavior often has implicit conno-
tations of maleness.
Critical theory also argues that there is implicit bias in what is chosen for
study. The powerful are studied, the powerless —in terms of class, race, or gen-
der—are frequently neglected. If men and women are differentially present in
senior-level positions, for instance, studies of powerful managers will not have
many women in the data. Sheriff and Campbell wrote, "Men have produced a
body of organizational theory and research in which men are assumed to be the
primary objects of study" (1992, p. 31).
Critical theory has a more explicit focus on social change than most social sci-
ence. Nothing is to be taken for granted, and the goal of research is to help the
i88 New Directions for Organization 'Theory
Three major themes recur in this overview of the field of organization studies
and the concepts and controversies that characterize attempts to understand these
ubiquitous social entities, first, there are subjects given comparatively short shrift
by the field even though they are important in the environment and pose some
interesting challenges for some strands of organization theory. Among such sub-
jects are the changing natures of organization boundaries, the employment rela-
tion, and the size distribution of organizations, changes in managerial autonomy
produced in part by changes in the capital markets, changing organizational de-
mography as the work force composition changes and as organizations interna-
tionalize, and changes in the degree of wage inequality in organizations and, as a
consequence, in the United States. There are also numerous subject domains in
which important models of behavior make competing predictions and in which
adherents of the various models tend to talk past each other, ignore competing
points of view, and dismiss conflicting evidence, discussions about interdisciplin-
ary cross-fertilization to the contrary.
Second, there is the issue of the very diversity of the field, its increasing
differentiation, and the profusion of theories, concepts, and methods as the scope
and boundaries of organization studies expand to incorporate ideas from whatever
nearby disciplines seem to be useful at the time. And third, there is the idea that
both of these facets of the field — what is ignored and the theoretical fragmenta-
tion—are consequences of the particular evolution and location of organization
studies. In particular, because organization studies is located increasingly in busi-
ness schools, particularly U.S. business schools, economics has come to play a
larger role. In this interdisciplinary setting, fewer disciplinary anchors or forces
guide or direct research, contributing to the theoretical diversity that has come to
characterize the field. Also, by being housed in business schools, the managerial
concerns focused often at the individual or at most the firm level of analysis loom
189
190 New Directions for Organization Theory
large, and social policy and broader macrosocial issues such as wage inequality,
size distributions, and the governance and control of organizations receive com-
paratively less attention. The particular research context favors an individualistic,
careerist, and differentiated pattern:
This last chapter builds on the review of the field accomplished in the pre-
ceding chapters by taking stock of the field's progress and offering some thoughts
about useful future directions for research, paths to avoid, and ideas about the
task of building organization theory. The last time I did this, I concluded that
"the dominant perspective on action . . . has been the rational model of choice"
and that "the vast majority of the research has focused on the individual level of
analysis" (Pfcffer, 1982, p. 255). Because organizations are in fact "social entities
characterized by demographic, relational, and physical (material?) structures"
(Mirvis, 1981, p. 2), I argued that research on networks and social relations, orga-
nizational demography, and the physical design of organizations would be pro-
ductive and useful. The ensuing years have seen at least two of those recommen-
dations or predictions come to fruition. There has been extensive research on
both organizational demography (e.g., Stewman, 1988; Mitman, 1992) and social
networks (e.g., Burt, 1992). In both instances, there have been substantial ad-
vances in ideas, methods, and results.
By contrast, there has not been much attention given to physical design (e.g.,
Darley and Gilbert, 1985), and if anything there has been less research on this
subject since the 1980s. And, as noted in chapter 7, the subject of organizational
design has also diminished in the amount of research attention received. One of
the things this final chapter does is to renew the call for research on the effects
of the physical or built environment and to argue for a rcvitalization of research
on other aspects of design. This is part of a recommendation for the field to
become more phenomenon driven and to consider using issues of relevance as
one way of navigating the thicket of ideas and concepts. This chapter also pursues
another theme from the earlier review — a concern with the state of paradigm
proliferation—although now the argument is more readily made that the costs of
being unconcerned about the state of organization studies as a scientific activity
are closer at hand.
If judged by the formation of new journals and the expansion of existing ones,
the field of organization studies is doing well. Organization Science, Journal of
Management Inquiry, and Organization are only three of several journals that
New Directions for Organization 'Theory 191
have appeared since 1990. The Academy of Management Journal has recently
expanded by 50 percent in publication frequency (from four to six issues per year)
and has also expanded the size of each issue, and the Academy of Management
Review is larger as well. The Journal of Management has enlarged its editorial
board, scope, and ambitions. There is a vigorous and growing set of outlets for
organizational research. Book publishing is also flourishing. More and more pub-
lishers have entered the field of business publishing, which includes publication
of work on management and organizations, and the number of titles published
has expanded virtually every year. Attendance at professional meetings is up sub-
stantially, as is the size and length of the program and membership in the Acad-
emy of Management.
If judged by the passion aroused by debates over whether or not dispositions
are useful for understanding organizational behavior (House et al., 1996), whether
a critical perspective on both organizations and organization theory is useful (Fer-
guson, 1984), whether the field should seek more paradigmatic unity (Van
Maancn, 1995b), and whether the field is too functionalist (e.g., Martin, 1992b),
intellectual activity in the field is vigorous. Indeed, the vigor and scope of organi-
zation studies make the task of providing an overview of the field daunting.
By some other indicators, however, a less sanguine view of the field emerges.
Hambrick (1994) bemoaned organization studies' lack of influence on the world
of practice and its corresponding lack of disciplinary power and attention. There
is, for instance, a Council of Economic Advisors but no Council of Organiza-
tional (or Management) Advisors, even though both government and social policy
could benefit from the insights and research of organization studies. Although
economic statistics are gathered regularly by official agencies, statistics that would
assist in studying organizations must still, for the most part, be gathered by indi-
vidual scholars rather than by agencies with data gathering as their mission. The
1990s have witnessed efforts to sharply curtail federally funded research in the
sciences generally, but it is the social sciences that are particularly hard hit as
medical research, for instance, has been more successful in defending itself.
There is concern with the pace of the development of organizational knowl-
edge and even if there is such progress in our understanding. Webster and Star-
buck (1988) argued that knowledge was developing slowly, arid they buttressed
this claim by noting that the strength of relationships, as measured by the size of
reported correlations on a variety of topics, was actually getting smaller rather
than larger over time. Miner (1984) examined the connection between the use-
fulness, scientific validity, and frequency of mention by scholars for twenty-four
organization theories. He found little correlation among the three measures. For
example, more useful or more valid theories were no more likely to be frequently
mentioned than less useful or valid ones. This disconnection between validity,
applicability, and prominence is troubling for a field that aspires to be scientifi-
cally valid and useful for understanding and managing organizations. Barley and
Kunda were concerned about the field's inability to get beyond an "Anglo-
American vision of social order [that] rests on an opposition between mechanistic
and organic solidarity, associated, respectively, with normative and rational ideolo-
gies of control" (1992, p. 386).
icj2 New Directions for Organization Theory
There are a number of things that might be done to help knowledge develop
and to make the field more useful and usable by managers and poliey makers.
Some of these are suggested in what follows, but the discussion does not purport
to be comprehensive. Addressing these issues requires the discussion and attention
of all of those interested in organizations and their study.
As noted in the first chapter, there is evidence that organi/ation studies is increas-
ingly influenced by and preoccupied with an economic orientation or model of
behavior. The study of organizations seems to be not so much bothered by physics
envy (Van Maanen, 1995b) as it is by economics envy, which at the end of the
day is potentially much more troublesome. After all, most of us don't really bor-
row physics theory too often to understand organizations (chaos theory notwith-
standing), but the temptation to rely on rational economic models of choice and
organizational behavior, regardless of their ability to develop novel insights or to
produce empirically valid predictions, remains compelling. In their analysis of
waves of managerial rhetoric, Barley and Kuncla argued that the normative em-
phasis of the writing about organizational culture would be supplanted by a "re-
surgence of rationalism" (1992, p. 394). This prediction has turned out to be true
and the trend was already well advanced at the time of their article.
Why should we care about this trend? Canella and Paetzold argued that we
should not worry about the growing dominance of economic models in organiza-
tion analysis: "Should members of another field wish to mount a hostile takeover
of organizational science, we say 'let them come,' . . . recognizing that such a
takeover requires the convincing of members of our field . . . that some particu-
lar perspective is the best one offered to date" (1994, p. 337).
But this argument is wrong on two counts. It is misleading in the first in-
stance for its implications about the politics of knowledge. Those advocating an
economic approach to organizational analysis (e.g., Barney and Ouchi, 1986; Hes-
terly, Liebeskind, and Zenger, 1990) have displayed a willingness —indeed, an
eagerness —to take on topics traditionally left to psychologists, sociologists, and
organizations scholars, topics, that is, such as organizational structure (Milgrom
and Roberts, 1992) and the structuring and effects of rewards and incentives
(Zenger, 1992). Many of those pushing an economic model are explicit in their
view that this mode of analysis is intended to replace, rather than to supplement,
other forms of organizational analysis. Thus, for instance, Jensen predicted "an
impending revolution in organization theory" (1983, p. 324) based on emerging
theoretical work in economics. Hirschleifer was even more direct: "The social
sciences can be regarded as in a process of coalescing. As economies 'impcrialisti-
cally' employs its tools of analysis over a wider range of social issues, it will become
sociology and anthropology and political science. . . . As these other disciplines
grow increasingly rigorous, they will not merely resemble but will he economics"
(1977, pp. 3-4, emphasis in original).
Canella and Paetzold recognized that "science is not a magnificent march
New Directions for Organization Theory 193
toward absolute truth, but a social struggle among the scholars of the profession
to construct truth" (1994, p. 332). But the implication of this fact is that truth
will not necessarily triumph since truth is itself socially determined. Or, the best
perspective offered to date is determined not by some unambiguous empirical test
but rather by which theory or perspective can muster the most social support for
its definition of truth and organizational reality. Consequently, those factors that
produce advantage in a social struggle will be important in determining which
theoretical perspective triumphs, and unity in perspective and consensus are
among the factors that provide the economic perspective with an edge in the
contest among models and theories.
Being unconcerned about the proliferation of economic explanations of orga-
nizations is also wrong as a matter of science. Chapter 3 detailed just a portion
of what is wrong with the economic model as a way of analyzing organizations.
Choshal and Moran (1996) have made the argument that not only are economic
models often theoretically and empirically misleading; they are also "bad for prac-
tice" and, by inference, a poor foundation on which to build public or manage-
ment policy. "Organization economics . . . sees managers as untrustworthy and
as requiring controls from supcrordinate levels in the organization" (Donaldson,
1995, p. 4), thereby contributing to a control-oriented set of management prac-
tices (e.g., the emphasis on hierarchy, incentives, and surveillance) that are both
expensive and counterproductive. If this is true — and certainly more research and
argument may be warranted —then there are important consequences for organi-
zations themselves from the field of organization studies being seduced by eco-
nomic logic and models. Recall that theories of behavior, to the extent they guide
policies and practices, can become true through their very implementation
(Frank, 1988; Pfeffer, 1994, chap. 4) because of the self-fulfilling, interconnected
nature of behavior. Organizational scholars have an obligation to develop and test
alternative models of behavior and not simply cede the field to economics.
Somewhat paradoxically, even as the various strands of organization theory
may have become increasingly diverse and disconnected, the field as a whole has
become more connected to adjacent social sciences that advocate the rational
actor model of behavior. Even as debate about positivism has occurred within the
field of organization studies, and even as critical perspectives have evolved, the
field has come to be increasingly dominated by models and methods that would
seem to epitomize what critical theory eschews. This is not to suggest that there
aren't subcultures of organization theory that embrace the humanities, postmod-
ernism, and critical theory or that there aren't those who seek to maintain a
diversity of paradigms. But the current trend seems clear.
One cause for the flirtation of the field with economic models is the absence of
paradigmatic consensus over problems, theories, or research methods. This ab-
sence of a well-developed scientific paradigm makes experimenting with anything
194 New Directions for Organization Theory
new or different more desirable. It also has the undesirable property of permitting
taste, virtually unconstrained by scientific norms and standards, to run rampant.
Two experienced organizations scholars with extensive experience as editors and
reviewers made this point eloquently: "Even when a paper contains a well-
articulated theory that fits the data, editors or reviewers may reject it or insist the
theory be replaced simply because it clashes with their particular conceptual
tastes" (Sutton and Staw, 1995, p. 372).
Organization studies, because of its lack of a strong anchor either in a disci-
pline or in phenomena, is susceptible to being caught up by fads and fashion.
Making this same observation, Zald wrote, "organizational studies follows the rat-
ings, responding not only to academic fads but to the whims and foibles of
academic hucksters arid the problem definitions of corporate executives" (1993,
p. 514). Eccles and Nohria noted that management theory was obsessed with new-
ness, even though the "new" principles were essentially the same as the old ones:
"Even a casual reading of the management literature over the past seventy-five
years, both popular and academic, shows that every age 'discovers' these principles
anew. . . . Even as the fundamental themes of management remain the same,
the words used to express them constantly portray them as new" (1992, p. 5).
Following fads and fashion without some sense of a theoretical research
agenda that can withstand the blandishments of novelty and uniqueness can
cause problems for the development of the field of study, as Barley and Kunda
noted: "to the degree that trends in organizational theory mirror trends in mana-
gerial discourse at large, our efforts may be neither cumulative nor pathbreaking.
Instead, in the long run, the tenor of our theorizing may amount to little more
than the turning of a small cog within a larger socioeconomic clock over which
no one has control" (1992, p. 394).
Absent some consensus on the goals of the scientific enterprise, the tempta-
tion to be caught up in the new and to pursue "what's interesting" has a greater
influence on the research process. The study of organizations would probably be
well served to spend more time confronting data and phenomena and less time
being fascinated by the new, the elegant, or the obscure for their own sake. Don-
aldson maintained that "the present degree of pluralism in US organization the-
ory is excessive and harbours severe problems of incoherence, lack of cumulation,
cynicism, faddism and despair that anything of lasting worth can be accom-
plished" (1995, p. 6). Commenting on the prevalence of the uniqueness value —
"a prescription that uniqueness is good, and therefore organization scholars
should attempt to make unique contributions to their discipline," Mone and Mc-
Kinley wrote:
M. S. Davis (1971) pointed out that the criterion for evaluating the importance
of a social science theory is typically iconoclasm, rather than the theory's truth
value. . . . Astley (1985) . . . argued that "the valne attached to a theory, and
the assessment of its contribution to scientific progress, is primarily determined
. . . by its intellectual novelty." . . . Staw (1985) advocated a role for journals as
generators of variety rather than selection-retention mechanisms. . . . Gioia and
Pitre (1990) suggested that organizational science should strive to be original,
and Martin (1992[bj) stated that we should "invent our own experiments, uncon-
New Directions for Organization Theory 195
strained by the usual ways of thinking and writing about empirical work." (1993,
pp. 285; 286)
This value of uniqueness and innovation for its own sake, not for its truth
value or for its ability to advance our understanding of organizations, is very much
at variance with the use of a theory-building strategy employing the logic of strong
inference. MacKenzie and House advocated a strong inference approach as a
paradigm development strategy for organization studies. They described strong
inference research as:
based on the assumption that: (a) all theories, no matter how good at explaining
a set of phenomena, are ultimately incorrect and consequently will undergo
modification over time . . . and (b) the fate of the better theories is to become
explanations that hold for some phenomena in some limited conditions. . . . A
guiding principle of the strong inference research strategy is the Popperian posi-
tion that pursuit of knowledge is more efficient when scientists deliberately set
out to disprove theories, or seek rejections, than when they attempt to assemble
proof for theories. (1978, p. 13)
The role of theory in the organization sciences is problematic. First of all, there
are important disagreements about what theory is or should be. Van Maanen
argued that "theory makes its way in intellectual communities (and beyond) as
much by its style as by its theoretical claims, logic, sweep, empirical evidence,
or methodological strictures" (1995a, p. 690). To some, rhetoric and theory are
inseparable and almost indistinguishable (Van Maanen, 1995a; 1995b). 'There is
even the question as to whether theory is useful. "Some prominent researchers
have argued the case against theory. . . . The field first needs more descriptive
narratives about organizational life" (Sutton and Staw, 1995, p. 378). Rich de-
scription, writing that is at times indistinguishable from literature or reporting, is
necessary to build strong theory over time (Van Maanen, 1989). Others question
whether the demands placed on scholars for both good theory and rigorous data
are compatible or even feasible. Sutton and Staw (1995, p. 380—381) noted: "On
the one hand, editors and reviewers plead for creative and interesting ideas. . . .
On the other hand, authors are skewered for apparent mismatches between their
theory and data. . . . Contradictory demands for both strong theory and precise
measurement are often satisfied only by hypocritical writing."
On the other hand, some organization scholars argue for theory development
and theory testing based on principles found in the natural as well as the social
sciences. Sacks, one of the founders of conversation analysis, a branch of ethno-
methodology, itself a qualitative branch of sociology, argued for the importance
of methods that were replicable and that could be transferred to others: "When I
started to do research . . . I figured that sociology could not be an actual science
unless it was able to handle the details of actual events, handle them formally,
and in the first instance, be informative about them in the direct ways in which
primitive sciences tend to be informative, that is, that anyone else can go and see
whether what was said is so" (1984, p. 26).
New Directions for Organization Theory 197
MacKenzie and House (1978) argued for the importance of not only devel-
oping theory but rejecting it when it did not fit the data, a method for cumulating
knowledge and over time refining theories so that they were increasingly rigorous
and applicable. Donaldson (1995) has argued not only for theory in general but
for a specific theory, structural contingency theory, and has made the broader
argument that cumulation and integration of insights and findings are important
for the advance of organization studies.
Given these disagreements and the general absence of paradigmatic consen-
sus, prospects for theory development may not be bright. Sutton and Staw wrote,
"Lack of consensus on exactly what theory is may explain why it is so difficult to
develop strong theory in the behavioral sciences" (1995, p. 372). In a similar vein,
Weick noted "how hard it is in a low-paradigm field, in which people are novice
theorists, to spot which of their efforts are theory and which arc not" (1995, p.
387).
One way to navigate the morass of different theories, and one way to evaluate
the theories that present themselves, is to ask the extent to which each theory or
perspective is linked to a particular, important organizational phenomenon and
affords insight into understanding that phenomenon. It is, in fact, possible to
make the case that many of the prominent organization theories did, in fact,
derive from a concern with specific, interesting phenomena. What happened,
however, was that over time the theories became objects of discourse and develop-
ment in and of themselves, often losing connection to their original foundations.
For instance, one important source of institutional theory was the observation
that organizational goals and their associated technology were often only loosely
coupled with actual organizational arrangements, including structures (J. W.
Meyer and Scott, 1983). Another observation was the effect on the adoption of
various practices of the mere frequency of adoption of similar practices in the
environment rather than some technical or efficiency rationale for their use (e.g.,
Tolbcrt and Zucker, 1983). Much of the interest in social networks and the effects
of the social environment on behavior came from the observation that organiza-
tional behavior was contagious, in a social network sense (e.g., G. F. Davis, 1991),
and from the interest in the sources of the differences in individual career attain-
ments and success (e.g., Granovetter, 1974). Resource dependence theory arose
from observations of interorganizational influence (e.g., Pfeffer, 1972a). And the
rationalizing model, with its concomitant effect on the analysis of organizational
incentives and their consequences, derived in part from the observation that peo-
ple often developed attitudes to be consistent with their behavior rather than the
reverse (Aronson, 1972; Staw, 1980) and that it was the case that rewards often
had perverse effects (A. Kohn, 1993).
Theoretical perspectives that have been less phenomenon driven have some-
times gotten themselves into trouble as a consequence. Economics, for example,
often starts with a phenomenon —for example, the spread of the multidivisional
structure, the prevalence of seniority-graded wage profiles—but doesn't investigate
the cause of the phenomenon because it assumes the cause: efficiency. Rather, it
models how an efficiency-based explanation, consistent with what is observed, can
be derived. Because observations, in this mode of theorizing, can never be used
198 New Directions for Organization Theory
to inform or develop the theory—rather, the theory is used to account for the
observations —there cannot be any refutation or specifying of the scope conditions
of the theory, and, as a consequence, there is also no development of the theory.
Structural contingency theory (Donaldson, 1995) suffered from a different prob-
lem. Although it began with a phenomenon —differences in organizational struc-
ture and what might affect those differences and result from them—-it fairly
quickly moved to conceptualize the phenomenon using dimensions that were
both far removed from practice and managerial conversation and also too com-
plex and convoluted to be explained or recalled even to a scientific audience.
In that sense, the theory departed from the original thing it began to explain —
understanding organizational structures in the service of developing the science
and practice of organizational design.
But design remains an important concern in organizations, and issues of de-
sign have much to recommend them as important ways of analyzing and under-
standing organizations. Pursuing a design perspective, broadly defined, would,
therefore, be useful in advancing organizational analysis.
Hatch's interpretation was that a certain amount of privacy and personal space
was useful for building relationships and holding meetings and that an absence
of privacy actually reduced interaction. Her study also found that task interdepen-
dence and task uncertainty were not related to self-reported interaction activity.
The variables that were the primary focus of her study, the characteristics of physi-
cal space, accounted for a sizable portion of the variation in seven categories of
self-reported work activity and in most instances explained more of the variance
than either individual demographics, individual sociability scores, or the firm or
type of job.
There is a second important way in which the physical environment affects
organizational behavior. R. A. Baron has summarized the extensive literature on
the influence of affect, "the mild, temporary shifts in current mood that virtually
everyone experiences" (1993, p. 64), on a number of outcomes including negotia-
tion, conflict, performance appraisals, task performance, and absenteeism and
turnover. In a series of imaginative studies, Baron has demonstrated that "social
factors, although of obvious importance, are clearly not the only source of positive
and negative affect. Another set of factors that can produce shifts in current
moods relates to aspects of the physical environment. Included here are such vari-
ables as ambient temperature, humidity, noise, crowding, air quality, lighting, and
other features of the physical surroundings in which people work or interact" (p.
74). He reported effects of temperature on aggression, of pleasant scents on goal
setting and self-confidence, and of lighting on a number of dimensions of work
performance.
It is striking that organizations have undergone a number of transformations
without the literature linking these changes very explicitly to considerations of the
built environment. Two such changes come to mind. First, there is an increasing
emphasis on working in teams and teamwork in organizations, particularly as the
ranks of middle management have been decimated. A physical environment built
to facilitate and encourage teamwork is one that looks different from one built to
enhance managerial status or to promote individual, contemplative decision mak-
ing. There are examples of redesigning built environments to promote a sense of
team identification—for instance, when Volvo moved to a more team-based
model of production in its Kalmar truck factory, it redesigned the layout to incor-
porate areas for each team that enhanced visual contact and promoted a sense of
team identification. But, we have yet to fully understand what a team-based physi-
cal setting would encompass, and studies that relate the team orientation of the
physical environment to the success of self-managed team interventions remain
to be done.
Second, people now more frequently work off-site, in arrangements that en-
tail telecommuting from home or that involve working at customer premises or
New Directions for Organization Theory 201
on the road. Because physical space is expensive, some organisations have taken
these new work arrangements as an opportunity to reduce their space require-
ments and thereby cut occupancy costs. The idea is to provide individuals with
space in a central office only when and as needed—an arrangement sometimes
called "hoteling." Permanent storage with temporary work spaces reduces the
need to provide offices that sit vacant while individuals are working off-site. Al-
though this arrangement obviously saves space, it sends an interesting message
about the importance of interacting with one's peers and colleagues and, indeed,
about one's identification with the organization. Again, studies about the effects
of this arrangement on organizational identification, commitment and turnover,
career advancement (which often requires sponsorship and the building of net-
works), and learning and idea sharing would be useful. As Darley and Gilbert
noted, "layout serves not only to regulate communication but to communicate
something of its own. . . . Interior layout serves as a signaling system that defines
role, rights, and responsibilities" (1985, p. 972).
The neglect of physical design in the literature is related to the current ne-
glect of organizational design. As one way of demonstrating this current neglect,
consider the literature cited by one of the proponents of a design perspective. In
his overview of the structural contingency literature, Donaldson (1995, p. 14)
cited some sixty-one studies. Almost half of them were from the 1970s (thirty) and
only two were from the 1990s. It is fair to state that the late 1960s and early 1970s
were the heyday of academic interest in both structural contingency theory and
organizational design. Although management consultants and practicing manag-
ers remain interested in questions of design, the research literature has waned.
The problem is that both physical and organizational design share an "engi-
neering" orientation to the field of organizations. Although the study of manage-
ment has important historical roots in engineering, engineering is out, at least
among many academics today. Even as consultants have come to play a larger
role in producing books about organizations and organizational processes, and
even as consulting firms and similar organizations are increasingly involved in
doing studies and research about organizational issues, the response from re-
searchers in academic settings seems to be profound ambivalence. Although
many academics participate in these outside activities, within the academy there
seems to be some dominant notion of theoretical purity. There seems to be an
implicit belief that the way to retain both status and position is to be caiculativcly
irrelevant. As a strategy of inquiry, it is not likely to work. As MacKenzie and
House argued, "much is to be learned from applying the knowledge gained in
the laboratory to outside settings. Application is not the sole justification for basic
research, but it is certainly helpful in generating counter examples to a theory"
(1978, p. 21). Their position, with which I concur, is that it is through "engi-
neering" activities that one learns the extent to which a theory is useful and valid
and hence acquires the empirical insight necessary to develop and revise theory.
Thus, there is a case to be made that organization studies would be well
served to be both more phenomena driven and more concerned with the applica-
bility of its ideas. In pursuing these goals, an emphasis on design seems war-
ranted. The issue is not just that good theory is practical, to rephrase Lewin's old
2O2 New Directions for Organization Theory
dictum. It is that in the welter of ideas that characterizes the field, the "engi-
neering" application of those ideas provides one way of distinguishing those that
help us understand organizations and those that don't.
Conclusion
Abegglen, ]. C., and G. Stalk, Jr. (1985). Kaisha: The Japanese corporation. New York:
Basic Books.
Abowd, J. M. (1990). Does performance-based managerial compensation affect corporate
performance? Industrial and ],abor Relations Review, 43, 52-73-S.
Abrahamson, E. (1996). Management fashion. Academy of Management Review, 21, 254-
285.
Achen, C. H. (1986). The statistical analysis of quasi-experiments. Berkeley: University of
California Press.
Acker, J. (1990). Hierarchies, jobs, bodies: A theory of gendered organizations. Gender and
Society, 4, 139-158.
Acker, J. (1992). Gendering organi/ational theory. In A. J. Mills and P. Tancred (Eds.),
Gendering organizational analysis, 248-260. Newbury Park, Calif: Sage.
Adler, P. S. (1992). The learning bureaucracy: New United Motors Manufacturing, Inc.
In B. M. Staw and I.. L. Cummings (Eds.), Research in organizational behavior:
Vol. 15, 111-194. Greenwich, Conn.: JAI Press.
Adler, P. S., and B. Borys (1989). Automation and skill: Three generations of research on
the NC case. Politics and Society, 17, 378-383.
Aiello, J. R. (1993). Electronic performance monitoring. Journal of Applied Social Psychol-
ogy, 23, W9-5Q7.
Aicllo, J. R., and A. Chomiak (1992). The effects of computer monitoring and distraction
on task performance. Unpublished ins., Rutgers University.
Aiello, }. R., and C. M. Svec (1993). Computer monitoring and work performance: Ex-
tending the social facilitation framework to electronic presence. Journal of Applied
Social Psychology, 23, 537-548.
Aldrich, H. E., and E. R. Auster (1986). Even dwarfs started small: Eiabilities of age and
si/c and their strategic implications. In B. M. Staw and L. L. Cummings (Eds.),
Research in organizational behavior: Vol. 8, 165-198. Greenwich, Conn.: }A1 Press.
Alexander, C. N., and G. W. Knight (1971). Situated identities and social psychological
experimentation. Sociometry, 34, 65-82.
205
206 References
Alexander, ]., B. Nuchols, J. Bloom, and S. Lcc (1995). Organizational demography and
turnover: An examination of multiform and nonlinear heterogeneity. Human Rela-
tions, 48, 1455-1480.
Alexander, V. D., and P. A. Thoits (1985). Token achievement: An examination of propor-
tional representation and performance outcomes. Social Forces, 64, 332-340.
Allen, M. P. (1974). The structure of interorganizational elite co-optation: Interlocking
corporate directorates. American Sociological Review, 39, 393-406.
Allen, R. W., D. L. Madison, I,. W. Porter, P. A. Renwick, and B. T. Mayes (1979).
Organizational politics: Tactics and characteristics of its actors. California Manage-
ment Review, 22, 77—83.
Allen, S. A. (1978). Organizational choices and general management influence networks
in divisionalized companies. Academy of Management Journal, 21, 341—365.
Allmendinger, J., and J. R. Mackman (1995). The more the better? A four-nation study of
the inclusion of women in symphony orchestras. Social Forces, 74, 423-460.
Alvesson, M., and H. Willmott (1992). On the idea of emancipation in management and
organization studies. Academy of Management Review, 17, 432-464.
Amabilc, T. M. (1985). Motivation and creativity: Effects of motivational orientation on
creative writers. Journal of Personality and Social Psychology, 48, 393-399.
Amabile, T. M., B. A. Hennessey, and B. S. Grossman (1986). Social influences on creativ-
ity: The effects of contracted-for reward. Journal of Personality and Social Psychol-
ogy, 50, 14-23.
Amburgey, T. L., D. Kelly, and W. P. Barnett (1993). Resetting the clock: The dynamics
of organizational failure. Administrative Science Quarterly, 38, 51—73.
American Management Association (1994). J994 AMA survey on downsizing: Summary of
key findings. New York: American Management Association.
Amick, B. C., and M. J. Smith (1992). Stress, computer-based work monitoring and mea-
surement systems: A conceptual overview. Applied Ergonomics, 23, 6-16.
Applebaum, E. (1987). Restructuring work: Temporary, part-time, and at-home employ-
ment. In II. I. Ilartmann (Ed.), Computer chips and paper clips: Technology and
women's employment, 268-310. Washington, D.C.: National Academy Press.
Archibald, W. P. (1974). Alternative explanations for self-fulfilling prophecy. Psychological
Bulletin, 81, 74-84.
Argyris, C. (1972). The applicability of organizational sociology. London: Cambridge Uni-
versity Press.
Argyris, C. (1995). On actionable knowledge. Unpublished ms. Harvard University, Gradu-
ate School of Business Administration.
Arkes, H. R., and C. Blumer (1985). The psychology of sunk cost. Organizational Behavior
and Human Decision Processes, 35, 124-140.
Armour, H. O., and D. J. Teecc (1978). Organizational structure and economic perfor-
mance: A test of the multidivisional hypothesis. Bell Journal of Economics, 9, 106-
122.
Aronson, E. (1972). The social animal. San Francisco: W. II. Freeman.
Aronson, E., and J. Mills (1959). The effect of severity of initiation on liking for a group.
journal of Abnormal and Social Psychology, 59, 177-181.
Arrow, K. (1982). Risk perception in psychology and economics. Economic Inquiry, 20, 1-
9.
Arthur, J. B. (1992). The link between business strategy and industrial relations systems in
American steel minimills. Industrial and Labor Relations Review, 45, 488-506.
Arthur, j. B. (1994). Effects of human resource systems on manufacturing performance
and turnover. Academy of Management Journal, 37, 670-687.
References 207
Barnett, W. P., and G. R. Carroll (1987). Competition and mutualism among early tele-
phone companies. Administrative Science Quarterly, 32, 400-421.
Barnett, W. P., and G. R. Carroll (1995). Modeling internal organizational change. Annual
Review of Sociology, 21, 217-236.
Barnett, W. P., H. R. Greve, and D. Y. Park (1994). An evolutionary model of organiza-
tional performance. Strategic Management Journal, 15, 11-28.
Barney, J. B., and W. G. Ouchi (Kds.) (1986). Organizational economics: Toward a new
paradigm for understanding and studying organizations. San Francisco: Jossey-Bass.
Baron, J. N. (1988). The employment relation as a social relation, journal of the Japanese
and International Economies, 2, 492-525.
Baron, J. N., I1'. R. Dobbin, and P. D. Jennings (1986). War and peace: The evolution of
modern personnel administration in U.S. industry. American Journal of Sociology,
92, 350-383.
Baron, ]. N., B. S. Mittman, and A. li. Newman (1991). Targets of opportunity: Organiza-
tional and environmental determinants of gender integration within the California
civil service, 1979-1985. American Journal of Sociology, 96, 1362-1401.
Baron, }. N., and A. I1',. Newman (1990). For what it's worth: Organizational and occupa-
tional factors affecting the value of work done by women and non-whites. American
Sociological Review, 55, 155-175.
Baron, ]. N., and J. Pfeffer (1994). The social psychology of organizations and inequality.
Social Psychology Quarterly, 57, 190-209.
Baron, R. A. (1993). Affect and organizational behavior: When and why feeling good (or
bad) matters. In J. K. Murnighan (Ed.), Social psychology in organizations: Ad-
vances in theory and research, 63-88.
Barrick, M. R., and M. K. Mount (1991). The Big Five personality dimensions and job
performance: A meta-aualysis. Personnel Psychology, 44, 1-26.
Barrick, VI. R., and M. K. Mount (1993). Autonomy as a moderator of the relationship
between the big five personality dimensions and job performance. Journal of Ap-
plied Psychology, 78, 111-118. '
Barry, B. (1978). Sociologists, economists, and democracy. Chicago: University of Chicago
Press.
Bass, B. M. (1985). Leadership and performance beyond expectations. New York: Free Press.
Baum, ). A. O. (1995). The changing basis of competition in organizational populations:
'I'he Manhattan hotel industry, 1898-1990. Social Forces, 74, 177-205.
Baum, J. A. C., and S. J. Me/ias (1992). Focalized competition and organizational failure
in the Manhattan hotel industry, 1898-1990. Administrative Science Quarterly, 37,
580-604.
Baum, J. A. C., and C. Oliver (1991). Institutional linkages and organizational mortality.
Administrative Science Quarterly, 36, 187-218.
Baum, J. A. C., and C. Oliver (1992). Institutional embeddedness and the dynamics of
organizational populations. American Sociological Review, 57, 540-559.
Baum, J. A. C., and W. W. Powell (1995). Cultivating an institutional ecology of organiza-
tions: Comment on Hannan, Carroll, Dundon, and Torres. American Sociological
Review, 60, 529- -538.
Baum, J. A. C., and J. V. Singh (1994). Organizational niches and the dynamics of organi-
zational mortality. American Journal of Sociology, 100, 346—380.
Baurnol, W. (1959). Business behavior, value, and growth. New York: Macmillan.
Bazerman, M. IF, R. I. Bceknn, and V. D. Schoorman (1982). Performance evaluation in
a dynamic context: A laboratory study of the impact of prior commitment to the
ralee. Journal of Applied Psychology, 67, 873-876.
References 209
Blaug, M. (1980). The methodology of economics. New York: Cambridge University Press.
Blauner, R. (1964). Alienation and freedom: The factor)' worker and his industry. Chicago:
University of Chicago Press.
Bluen, S. D., ]. Barling, and W. Burns (1990). Predicting sales performance, job satisfac-
tion, and depression by using the Achievement Strivings and Impatience-Irritability
dimensions of Type A behavior. Journal of Applied Psychology, 75, 212-216.
Bockcr, W. (1989). The development and institutionalization of subunit power in organiza-
tions. Administrative Science Quarterly, 34, 388-410.
Boeker, W. (1992). Power and managerial dismissal: Scapegoating at the top. Administra-
tive Science Quarterly, 37, 400-421.
Boland, L. A. (1979). A critique of Friedman's critics. Journal of Economic Literature, 17,
503-522.
Borjas, C. (1980). Wage determination in the federal government: The role of constituents
and bureaucrats. Journal of Political Economy, 88, 1110-1147.
Borjas, G. (1982). The politics of employment discrimination in the federal bureaucracy.
journal of Law and Economics, 25, 271-299.
Borjas, G. (1983). The measurement of race and gender wage differentials: Evidence from
the federal sector. Industrial and Labor Relations Review, 37, 79-91.
Bouchard, T. ]., and M. McGue (1981). Familial studies of intelligence: A review. Science,
212, 1055-1059.
Bougon, M. G. (1983). Uncovering cognitive maps: The self-Q technique. In G. Morgan
(Ed.), Beyond method: Strategies for social research. Beverly Hills, Calif: Sage.
Bougon, M. G., K. E. Weick, and D. Binkhorst (1977). Cognition in organizations: An
analysis of the Utrecht Jazz Orchestra. Administrative Science Quarterly, 22, 606-
639.
Boulding, K. (1968). The organizational revolution: A study of the ethics of economic organi-
zation. Chicago: Quadrangle Books.
Bowman, E. H. (1963). Consistency and optimality in managerial decision-making. Man-
agement Science, 9, 310-321.
Boxman, E. A. W., P. M. De Graaf, and H. D. Flap (1991). The impact of social and
human capital on the income attainment of Dutch managers. Social Networks, 13,
51-73.
Brass, D. J. (1985). Men's and women's networks: A study of interaction patterns and
influence in an organization. Academy of Management Journal, 28, 327-343.
Brass, D. ]., and M. E. Burkhardt (1993). Potential power and power use: An investigation
of structure and behavior. Academy of Management Journal, 36, 441-470.
Bravcrman, II. (1974). Labor and monopoly capital. New York: Monthly Review Press.
Brchm, J. W. (1966). A theory of psychological reactance. New York: Academic Press.
Bresnen, M., and C. Fowler (1994). The organizational correlates and consequence of
subcontracting: Evidence from a survey of South Wales. Journal of Management
Studies, 31, 847-864.
Bridges, W. E, and W. J. Villemcz (1994). The employment relationship: Causes and conse-
quences of modem personnel administration. New York: Plenum.
Brief, A. P., A. I). Butcher, and L. Roberson (1995). Cookies, disposition, and job attitudes:
The effects of positive mood-inducing events and negative affectivity on job satisfac-
tion in a field experiment. Organizational Behavior and Human Decision Processes,
62, 55-62.
Brinkerhoff, R. O. (Ed.) (1989). Evaluating training programs in business and industry.
San Francisco: Jossey-Bass.
References 211
Brockner, JL, and J. Z. Rubin (1985). Entrapment in escalating conflicts. New York:
Springer-Verlag.
Brockner, J., J. Z. Rubin, and E. Lang (1981). Face saving and entrapment. Journal of
Experimental Social Psychology, 17, 68-79.
Brookes, M. J., and A. Kaplan (1972). The office environment: Space planning and af-
fective behavior. Human Factors, 14, 373-391.
Brown, C. (Ed.) (1994). The competitive semiconductor manufacturing human resources
project: First interim report. Berkeley: University of California, Berkeley, Institute of
Industrial Relations.
Budros, A. (1994). Analyzing unexpected density dependence effects on organizational
births in New York's life insurance industry. Organization Science, 5, 541-553.
Burawoy, M. (1979). Manufacturing consent. Chicago: University of Chicago Press.
Burawoy, M. (1985). The politics of production. New York: Verso.
Burns, L. R., and D. R. Wholey (1993). Adoption and abandonment of matrix manage-
ment programs: Effects of organizational characteristics and interorganizational net-
works. Academy of Management Journal, 36, 106-138.
Burns, T., and G. M. Stalker (1961). The management of innovation. London: lavistock.
Burrell, G., and G. Morgan (1979). Sociological paradigms and organizational analysis.
London: Heinemann.
Burt, R. S. (1980). Autonomy in a social topology. American Journal of Sociology, 85, 892-
925.
Burt, R. S. (1983). Corporate profits and co-optation. New York: Academic Press.
Burt, R. S. (1987). Social contagion and innovation: Cohesion versus structural equiva-
lence. American Journal of Sociology, 92, 1287-1335.
Burt, R. S. (1992). Structural holes: The social structure of competition. Cambridge, Mass.:
Harvard University Press.
Burt, R. S., K. P. Christman, and H. C. Kilburn, Jr. (1980). Testing a structural theory of
corporate co-optation: Interorganizational directorate ties as a strategy for avoiding
market constraints on profits. American Sociological Review, 45, 821-841.
Galas, M. (1993). Deconstructing charismatic leadership: Re-reading Weber from the
darker side. Leadership Quarterly, 4, 305-328.
Galas, M., and L. Smircich (1992). Using the f-word: Feminist theories and the social
consequences of organizational research. In A. J. Mills and P. Tancred (Eds.), Gen-
dering organizational analysis, 222-234. Ncwbury Park, Calif: Sage.
Calder, B. J. (1977). An attribution theory of leadership. In B. M. Staw and G. R. Salancik
(Eds.), New directions in organizational behavior, 179-204. Chicago: St. Glair Press.
Caldwell, D. F., C. A. O'Reilly III, and J. H. Morris (1983). Responses to an organizational
reward: A field test of the sufficiency of justification hypothesis. Journal of Applied
Psychology, 44, 506-514.
Cammann, C., E. E. Lawler III, G. E. Ledford, and S. E. Seashore (1984). Management-
labor cooperation in quality of worklife experiments: Comparative analysis of eight
cases. Report to the U.S. Department of Labor. Ann Arbor: Survey Research Cen-
ter, University of Michigan.
Campbell, J. P. (1990). An overview of the army selection and classification project (Proj-
ect A). Personnel Psychology, 43, 231-239.
Canella, A. A., Jr., and R. L. Paetzold (1994). Pfeffer's barriers to the advance of organiza-
tional science: A rejoinder. Academy of Management Review, 19, 331-341.
Cappelli, P. (1995). Rethinking employment. British Journal of Industrial Relations, 33,
563-602.
212 References
Carlson, R. C). (1972). School superintendents: Career and performance. Columbia, Ohio:
Merrill.
Carnevale, P., and A. Isen (1986). The influence of positive affect and visual access on the
discovery of integrative solutions in bilateral negotiation. Organizational Behavior
and Human Decision Processes, 37, 1-13.
Carroll, C. R. (1983). A stochastic model of organizational mortality: Review and reanaly-
sis. Social Science Research, 12, 303—329.
Carroll, G. R. (1984). Dynamics of publisher succession in newspaper organizations. Ad-
ministrative Science Quarterly, 29, 93-113.
Carroll, G. R. (1985). Concentration and specialization: Dynamics of niche width in popu-
lations of organizations. American journal of Sociology, 90, 1262-1283.
Carroll, G. R., and J. Delacroix (1982). Organizational mortality in the newspaper indus-
tries of Argentina and Ireland: An ecological approach. Administrative Science
Quarterly, 27, 169-198.
Carroll, G. R., and P. Huo (1986). Organizational task and institutional environments in
ecological perspective: Findings from the local newspaper industry. American Jour-
nal of Sociology, 91, 838-873.
Carroll, G. R., and A. Swaminarhan (1992). The organizational ecology of strategic groups
in the American brewing industry from 1975 to 1990. Industrial and Corporate
Change, 1, 65-97.
Carroll, C. R., and J. B. Wade (1991). Density dependence in the evolution of the Ameri-
can brewing industry across different levels of analysis. Social Science Research, 20,
271-302.
Carson, R. C. (1989). Personality. Annual Review of Psychology, 40, 227-248.
Castro, ). (1993). Disposable workers. Time, March 29, 1993, 43-47.
Chamberlain, K. (1988). On the structure of well-being. Social Indicators Research, 20,
581-604.
Chandler, A. (1962). Strategy and structure: Chapters in the history of the American indus-
trial enterprise. Cambridge, Mass.: MIT Press.
Channon, D. R (1973). 'lire strategy and structure of British enterprise. London: Mac-
rnillan.
Chatman, J. A. (1989). Improving organizational research: A model of person-organization
fit. Academy of Management Review, 14, 333-349.
Chatman, J. A. (1991). Managing people and organizations: Selection and socialization in
public accounting firms. Administrative Science Quarterly, 36, 459-484.
Chatman, ]. A., and S. G. Barsade (1995). Personality, organizational culture, and coopera-
tion: Evidence from a business simulation. Administrative Science Quarterly, 40,
423-443.
Chatman, J. A., D. R Caldwell, and C. A. O'Reilly (1995). Managerial personality and
early career success: A profile comparison approach. Unpublished ins., University
of California, Berkeley, Haas School of Business.
Chatman, J. A., and K. A. Jehn (1994). Assessing the relationship between industry charac-
teristics and organizational culture: How different can yon be? Academy of Manage-
ment Journal, 37, 522-553.
Chertkoff, J. M., and M. Conlcy (1967). Opening offer and frequency of concession as a
bargaining strategy. Journal of Personality and Social Psychology, 7, 181-185.
Child, J. (1972). Organization structure, environment, and performance: The role of strate-
gic choice. Sociology, 6, 1-22.
Child, J. (1973). Predicting and understanding organization structure. Administrative Sci-
ence Quarterly, 18, 168-185.
References 213
Child, J. (1975). Managerial and organizational factors associated with company perfor-
mance, part 2: A contingency analysis. Journal of Management Studies, 12, 12-27.
Child, J., and R. Mansfield (1972). Technology, size, and organization structure. Sociology,
6, 369-393.
Christensen, K. (1995). Contingent work arrangements in family-sensitive corporations. Bos-
ton: Boston University, Center on Work and Family.
Cialdini, R. B. (1988). Influence: Science and practice. Glenview, 111.: Scott, Foresman.
Cialdini, R. B., N. Eisenberg, B. L. Green, K. Rhoads, and R. Bator (1994). Undermining
the undermining effect of reward on intrinsic interest: When unnecessary condi-
tions are sufficient. Unpublished ms. Arizona State University, Department of Psy-
chology.
Clawson, D. (1980). Bureaucracy and the labor process: The transformation of U.S. industry,
1860-1920. New York: Monthly Review Press.
Clawson, D., and A. Neustadtl (1989). Interlocks, PACs, and corporate conservatism. Amer-
ican lournal of Sociology, 94, 749-773.
Clawson, D., A. Neustadtl, and J. Bearden (1986). The logic of business unity: Corporate
contributions in the 1980 election. American Sociological Review, 51, 797-811.
Clegg, S. R. (1990). Modern organizations: Organization studies in the postmodern world.
London: Sage.
Coase, R. [I. (1937). The nature of the firm. Economica, 4, 386-405.
Coasc, R. H. (1978). Economics and contiguous disciplines. Journal of Legal Studies, 7,
201-211.
Coasc, R. H. (1983). The new institutional economics. Journal of Institutional and Theoret-
ical Economics, 1, 229-231.
Cohen, Y., and J. Pfeffer (1986). Organizational hiring standards. Administrative Science
Quarterly, 31, 1-24.
Coker, D. A., M. A. Neale, and G. B. Northcraft (1987). Structural and individual influ-
ence on the process and outcome of negotiation. Unpublished ms. University of
Arizona, College of Business.
Cole, R. K. (1979). Work, mobility, and participation: A comparative study of American
and Japanese industry. Los Angeles: University of California Press.
Cole, R. E. (1989). Strategies for learning. Berkeley: University of California Press.
Cole, S. (1983). The hierarchy of the sciences? American Journal of Sociology, 89, 111-
139.
Colcman, J. S. (1990). Foundations of social theory. Cambridge, Mass.: Harvard University--
Press.
Coleman, J. S., E. Katz, and H. Men/el (1966). Medical innovation. New York: Bobbs-
Merrill.
Collins, J. C., and J. I. Porras (1994). Built to last: Successful habits of visionary companies.
New York: IlarperBusiness.
Collins, R. (1974). Where are educational requirements for employment highest? Sociol-
ogy of Education, 47, 419-442.
Collins, R. (1979). The credential society: An historical sociology of education and stratifica-
tion. New York: Academic Press.
Conrath, C. W. (1973). Communication patterns, organizational structure, and man:
Some relationships. Human Factors, 15, 459—470.
Cook, K. S., R. M. Emerson, M. R. Gillmore, and T. Yamagishi (1983). The distribution
of power in exchange networks: Theory and experimental results. American Journal
of Sociology, 89, 275-305.
Costa, P. T, and R. R. McCrae (1988). Personality in adulthood: A six-year longitudinal
214 References
study of self-reports and spouse ratings on the NEO Personality Inventory, journal
of Personality and Social Psychology, 54, 853-863.
Cowherd, D. M., and D. I. Lcvine (1992). Product quality and pay equity between lower-
level employees and top management: An investigation of distributive justice the-
ory. Administrative Science Quarterly, 37, 302-320.
Cropanzano, R., and K. James (1990). Some methodological considerations for the behav-
ioral genetic analysis of work attitudes. Journal of Applied Psychology, 75, 433-439.
Cross, J. C. (1983). A theory of adaptive economic behavior. Cambridge: Cambridge Uni-
versity Press.
Crystal, G. S. (1991). In search of excess: The overcompensation of American executives.
New York: Norton.
Cyert, R. M., W. R. Dill, and J. G. March (1958). The role of expectations in business
decision making. Administrative Science Quarterly, 3, 307-340.
Cyert, R. M., and J. C. March (1963). A behavioral theory of the firm. Knglewood Cliffs,
N.J.: Prentice Hall.
Cyert, R. M., I I . A. Simon, and D. B. Trow (1956). Observation of a business decision.
Journal of Business, 29, 237-248.
Dahl, R. A. (1957). The concept of power. Behavioral Science, 2, 201-215.
Darley, J. M., and D. T. Gilbert (1985). Social psychological aspects of environmental
psychology. In C. Lindzey and E. Aronson (Eds.), Handbook of social psychology,
3rd Ed., 949-991. New York: Random House.
D'Aveni, R. A. (1990). Top managerial prestige and organizational bankruptcy. Organiza-
tion Science, J, 121-142.
Davis, G. F. (1991). Agents without principles? The spread of the poison pill through the
intercorporate network. Administrative Science Quarterly, 36, 583-613.
Davis, G. I1'., and T. A. Thompson (1994). A social movement perspective on corporate
control. Administrative Science Quarterly, 39, 141-173.
Davis, M. S. (1971). That's interesting! 'Ibwards a phenomenology of sociology and a soci-
ology of phenomenology. Philosophy of the Social Sciences, 1, 309-344.
Davis, S. (1984). Managing corporate culture. Cambridge, Mass.: Ballinger.
Davis-Blake, A., and J. Pfeffer (1989). Just a mirage: The search for dispositions! effects in
organizational research. Academy of Management Review, 14, 385—400.
Davis-Blake, A., and B. Uzzi (1993). Determinants of employment cxtemalization: A study
of temporary workers and independent contractors. Administrative Science Quar-
terly, 38, 195-223.
Deal, T., and A. Kennedy (1982). Corporate cultures: The rites and rituals of corporate life.
Reading, Mass.: Addison-Wesley.
Deci, E. (1975). Intrinsic motivation. New York: Plenum.
Demb, A., and F. I1'. Neubauer (1992). The corporate board: Confronting the paradoxes.
New York: Oxford University Press.
Dennison, D. (1990). Corporate culture and organizational effectiveness. New York: John
Wiley.
Deutsch, M., and II. Gerard (1955). A study of normative and informational social influ-
ences on individual judgment, journal of Abnormal and Social Psychology, 51, 629—
636.
Diener, E. (1990). Challenges in measuring subjective well-being and ill-being. Working
paper, University of Illinois, Urbana, Department of Psychology.
Digrnan, J. M. (1990). Personality structure: Emergence of the five-factor model. Annual
Review of Psychology, 41, 417-440.
Dillon, R. L. (1987). The changing labor market: Contingent workers and the self-
References 215
Edwards, R. C. (1979). Contested terrain: 'ihe transformation of the workplace in the twenti-
eth century. New York: Basic Books.
Ehrenberg, R. G. (1990). Comment. In A. S. Blinder (Ed.), Paying for productivity: A look
at the evidence, 88-94. Washington, D.C.: Brookings.
Ehrenberg, R. G., and M. L. Bognanno (1988). Do tournaments have incentive effects?
NBER Research Working Paper No. 2638. Washington, D.C.: National Bureau of
Economic Research.
Ehrenberg, R. G., and M. E. Bognanno (1990). The incentive effects of tournaments
revisited: Evidence from the European PGA tour. Industrial and Labor Relations
Review, 43, 74-S-88-S.
Eisenbergcr, R. (1992). Learned industriousness. Psychological Review, 99, 248-267.
Eisenhardt, K. (1989). Agency theory: An assessment and review. Academy of Management
Review, 14, 57-74.
Ely, R. J. (1994). The effects of organi/ational demographics and social identity on rela-
tionships among professional women. Administrative Science Quarterly, 39, 203-
238.
Ely, R. J. (1995). The power in demography: Women's social constructions of gender iden-
tity at work. Academy of Management Journal, 38, 589-634.
Emerson, R. M. (1962). Power-dependence relations. American Sociological Review, 27,
31-41.
Engelbrecht, A. S., and A. H. Fischer (1995). The managerial performance implications
of a developmental assessment center process. Human Relations, 48, 387—404.
England, P., G. Earkas, B. S. Kilbourne, and T. Don (1988). Explaining occupational sex
segregation and wages: Findings from a model with fixed effects. American Sociolog-
ical Review, 66, 147-168.
Enzle, M. E., and S. G. Anderson (1993). Snrveillant intentions and intrinsic motivation.
journal of Personality and Social Psychology, 64, 257-266.
Et/ioni, A. (1988). The moral dimension: Toward a new economics. New York: Free Press.
Evans, M. G., M. N. Kiggundn, and R. J. House (1979). A partial test and extension of
the job characteristics model of motivation. Organizational Behavior and Human
Performance, 24, 354-371.
Ferber, M. A., and B. Kordick (1978). Sex differentials in the earnings of Ph.D.s. Industrial
and Labor Relations Review, 35, 550-564.
Ferguson, K. (1984). The feminist case against bureaucracy. Philadelphia: Temple Univer-
sity Press.
Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117-
140.
Festinger, L. (1957). A theory of cognitive dissonance. Stanford, Calif.: Stanford University
Press.
Festinger, L., S. Schachter, and K. Back (1950). Social pressures in informal groups. New
York: I larper.
Finkelstein, S. (1992). Power in top management teams: Dimension, measurement, and
validation. Academy of Management Journal, 35, 505-538.
Finkelstein, S. (1996). Inter-industry merger patterns and resource dependence: A replica-
tion and extension of Pfeffer (1972). Unpublished working paper, Dartmouth Col-
lege, Tuck Graduate School of Business.
Finkelslein, S., and D. C. Hambrick (1990). Top-managemcnt-teani tenure and organiza-
tional outcomes: The moderating role of managerial discretion. Administrative Sci-
ence Quarterly, 35, 484-503.
References 217
Finkelstein, S., and D. C. Hambrick (1996). Strategic leadership: '['op executives and their
effects on organizations. St. Paul, Minn.: West.
Fischhoff, B. (1982). Debiasing. In D. Kahneman and A. Tversky (Eds.), Judgment under
uncertainty: Heuristics and biases. New York: Cambridge University Press.
Fligstein, N. (1985). The spread of the multidivisional form among large firms, 1919—
1979. American Sociological Review, 50, 377-391.
Fligstein, N. (1987). The intraorganizational power struggle: The rise of finance presidents
in large corporations, 1919-1979, American Sociological Review, 52, 44-58.
Fligstein, N. (1990). 'Hie transformation of corporate control. Cambridge, Mass.: Harvard
University Press.
Fligstein, N. (1991). The structural transformation of American industry: An institutional
account of the causes of diversification in the largest firms, 1919-1979. In W. W.
Powell and P. J. DiMaggio (Eds.), The new institiitionalism in organizational analy-
sis, 311-336. Chicago: University of Chicago Press.
Fligstein, N., and P. Brantley (1992). Bank control, owner control, or organi/ational dy-
namics: Who controls the large modern corporation? American Journal of Sociology,
98, 280-307.
Fox, M. F. (1985). Location, sex-typing, and salary among academics. Work and Occupa-
tions, 12, 186-205.
Frank, R. H. (1984). Are workers paid their marginal products? American Economic Review,
74, 549-571.
Frank, R. H. (1988). Passions within reason: The strategic role of the emotions. New York:
Norton.
Frank, R. H. (1990). A theory of moral sentiments. In J. ). Mansbridge (Ed.), Beyond self-
interest, 71-96. Chicago: University of Chicago Press.
Frank, R. II., T. Gilovich, and D. T. Regan (1993). Does studying economics inhibit
cooperation? journal of Economic Perspectives, 7, 159-171.
Freeman, J., G. R. Carroll, and M. T. Ilannan (1983). The liability of newness: Age
dependence in organizational death rates. American Sociological Review, 48, 692-
710.
Freeman, }., and M. T. Ilannan (1983). Niche width and the dynamics of organi/ational
populations. American Journal of Sociology, 88, 116-145.
Freeman, R. B., and J. L Medoff (1984). What do unions do? New York: Basic.
French, J. R. P., Jr., and B. Raven (1968). The bases of social power. In D. Cartwright and
A. Zander (Eds.), Group dynamics, 3rd ed., 259-269. New York: Harper and Row.
Friedman, M. (1953). The methodology of positive economics. In M. Friedman, Essays in
positive economics. Chicago: University of Chicago Press.
Fucini, J. J., and S. F'ucini (1990). Working for the Japanese. New York: Collier Macmillan.
Funder, D. C., and C. R. Colvin (1991). Explorations in behavioral consistency: Properties
of persons, situations, and behaviors. Journal of Personality and Social Psychology,
60, 773-794.
Gabor, A. (1990). The man who discovered quality. New York: Times Books.
Galanter, M. (1989). Cults: Faith, healing, and coercion. New York: Oxford University
Press.
Galaskiewicz, J., and R. S. Burt (1991). Interorgani/ation contagion in corporate philan-
thropy. Administrative Science Quarterly, 36, 88-105.
Galaskiewiez, J., and S. Wasserman (1989). Mimetic processes within an interorgani/a-
tional field: An empirical test. Administrative Science Quarterly, 34, 454-479.
Galaskiewicz, J., and S. Wasserman, B. Rauschenbach, W. Bielefeld, and P. Mullaney
218 References
(1985). The influence of corporate power, social status, and market position on
corporate interlocks in a regional network. Social Forces, 64, 403-431.
Galbraith, J. R. (1973). Designing complex organizations. Reading, Mass.: Addison-
Wcsley.
Garrison, VV. A., and N. R. Scotch (1964). Scapegoating in baseball. American Journal of
Sociology, 70, 69-76.
Gaud/, J., and V. V. Murray (1980). The experience of workplace politics. Academy of
Management Journal, 23, 237-251.
Gardner, J. W. (1990). On leadership. New York: Free Press.
Gargiulo, Martin (1993). Two-step leverage: Managing constraint in organi/ational politics.
Administrative Science Quarterly, 38, 1-19.
Garrahan, P., and P. Stewart (1992). 'The Nissan enigma: Flexibility at work in a local
economy. London: Manscll.
Garson, B. (1988). The electronic sweatshop. New York: Simon and Schuster.
Gavin, M. B., S. G. Green, and G. T. Fairhurst (1995). Managerial control strategics for
poor performance over time and the impact on subordinate reactions. Organiza-
tional Beliavior and Human Decision Processes, 63, 207-221.
Gaziel, IT. H. (1994). Managerial studies and perceived job performance: An Israeli ease
study. Public Personnel Management, 23, 341-356.
George, J. M. (1989). Mood and absence. Journal of Applied Psychology, 74, 317-324.
George, J. M. (1991). State or trait: Effects of positive mood on prosocial behaviors at
work. Journal of Applied Psychology, 76, 299-307.
George, J. M. (1992). The role of personality in organizational life: Issues and evidence.
Journal of Management, 18, 185-213.
Gcorgenson, D. L. (1982). The problem of transfer calls for partnership. Training and
Development Journal, 36, 75-78.
Gerhart, B. (1987). How important are dispositional factors as determinants of job satisfac-
tion? Implications for job design and other personnel programs. Journal of Applied
Psychology, 72, 366-373.
Gerlach, M. (1992). Alliance capitalism: The social organization of Japanese business.
Berkeley: University of California Press.
Ghoshal, S., and P. Moran (1996). Bad for practice: A critique of the transaction cost
theory. Academy of Management Review, 21, 13-47.
Giddens, A. (1982). Power, the dialectic of control, and class structuration. In A. Giddens
and G. Mackenzie (Fds.), Social class and the division of labour. Cambridge: Cam-
bridge University Press.
Gioia, D. A., and E. Pitre (1990). Multiparadigm perspectives on theory building. Academy
of Management Review, 15, 584-602.
Gioia, D. A., and 11. P. Sims, Jr. (1986). Introduction: Social cognition in organizations.
In H. P. Sims, Jr., and D. A. Gioia (Eds.), The thinking organization, 1-19. San
Francisco: Josscy-Bass.
Goldberg, L. R. (1990). An alternative "description of personality": The big-five factor
structure. Journal of Personality and Social Psychology, 59, 1216-1229.
Goldberg, V. P. (1980). Bridges over contested terrain: Exploring the radical account of
the employment relationship. Journal of Economic Behavior and Organization, 1,
249-274.
Goldman, P., and D. R. Van Houtcn (1977). Managerial strategies and the worker: A
Marxist analysis of bureaucracy. Sociological Quarterly, 18, 108—125.
Goodman, P. S. (1980). Realities of improving the quality of work life. Labor Law Journal,
31, 487-494.
References 219
Howell, J. M., and P. J. Frost (1989). A laboratory study of charismatic leadership. Organi-
zational Behavior and Human Decision Processes, 43, 243-269.
Huber, J. (1990). Macro-micro links in gender stratification. American Sociological Review,
55, 1-10.
Hunter, J. E. (1986). Cognitive ability, cognitive aptitudes, job knowledge, and job perfor-
mance. Journal of Vocational Behavior, 29, 340-362.
Huselid, M. A. (1995). The impact of human resource management practices on turnover,
productivity, and corporate financial performance. Academy of Management Jour-
nal, 38, 635-672.
Ibarra, M. (1992). Homophily and differential returns: Sex differences in network structures
and access in an advertising firm. Administrative Science Quarterly, 37, 422-447.
Ibarra, H. (1993a). Network centrality, power, and innovation involvement: Determinants
of technical and administrative roles. Academy of Management journal, 36, 471-
501.
Ibarra, H. (1993b). Personal networks of women and minorities in management: A concep-
tual framework. Academy of Management Review, 18, 56-87.
Ibarra, H. (1995). Race, opportunity, and diversity of social circles in managerial networks.
Academy of Management Journal, 38, 673-703.
Ibarra, H., and S. B. Andrews (1993). Power, social influence, and sense making: Effects of
network centrality and proximity on employee perceptions. Administrative Science
Quarterly, 38, 277-303.
Ichniowski, C. (1986). The effects of grievance activity on productivity. Industrial and
Labor Relations Review, 40, 75-89.
Ichniowski, C., K. Shaw, and G. Prennushi (1993). The effects of human resource man-
agement practices on productivity. Working paper. Columbia University, Graduate
School of Business Administration.
Irving, R. H., C. A. Higgins, and F. R. Safayeni (1986). Computeri/ed performance moni-
toring systems: Use and abuse. Communications of the ACM, 29, 794—801.
Isen, A. M., and P. F. Levin (1972). Effects of feeling good on helping: Cookies and
kindness, journal of Personality and Social Psychology, 21, 384-388.
Izraeli, D. N. (1983). Sex effects or structural effects? An empirical test of Kanter's theory
of proportions. Social Forces, 62, 153-165.
Jackson, P. B., P. A. Thoits, and H. F. Taylor (1995). Composition of the workplace and
psychological well-being: The effects of tokenism on America's black elite. Social
Forces, 74, 543-557.
Jackson, S. F,., J. F. Brett, V. I. Sessa, D. M. Cooper, J. A. Julin, and K. Peyronnin (1991).
Some differences make a difference: Individual dissimilarity and group heterogene-
ity as correlates of recruitment, promotions, and turnover. Journal of Applied Psy-
chology, 76, 675-689.
Jacobs, J. A. (1992). Women's entry into management: Trends in earnings, authority, and
values among salaried managers. Administrative Science Quarterly, 37, 282—301.
Jacoby, S. M. (1985). Employing bureaucracy. New York: Columbia University Press.
Jacoby, S. M. (1990). The new institutionalism: What can it learn from the old? Industrial
Relations, 29, 316-359.
Jensen, M. C. (1983). Organization theory and methodology. Accounting Review, 50, 319-
339.
Jensen, M. C., and W. H. Meckling (1976). Theory of the firm: Managerial behavior,
agency costs, and ownership structure. Journal of Financial Economics, 3, 305—360.
Jensen, M. C., and K. J. Murphy (1987). Are executives' compensation contracts structured
References 223
Kramer, R. M., and D. M. Messick (Eds.) (1995). Negotiation as a social process. Thou-
sand Oaks, Calif.: Sage.
Kramer, R. M., E. Newton, and P. L. Pomrnerenke (1993). Self-enhancement hiases and
negotiator judgment: Effects of self-esteem and mood. Organizational Behavior and
Human Decision Processes, 56, 110-133.
Kramer, R. M., P. L. Pommerenkc, and E. Newton (1993). The social context of negotia-
tion: Effects of social identity and accountability on negotiator judgment and deci-
sion making. Journal of Conflict Resolution, 37, 633-656.
Kravetz, D. J. (1988). The human resource revolution: Implementing progressive manage-
ment practices for bottom-line success. San Francisco: Josscy-Bass.
Kuhn, T. S. (1970). The structure of scientific revolutions, 2nd ed. Chicago: University of
Chicago Press.
Kunda, G. (1992). Engineering culture: Control and commitment in a high-tech corpora-
tion. Philadelphia: Temple University Press.
Lachman, R. (1989). Power from what? A reexamination of its relationships with structural
conditions. Administrative Science Quarterly, 34, 231-251.
Lambert, R. A., D. E. Larcker, and K. Weigelt (1993). The structure of organizational
incentives. Administrative Science Quarterly, 38, 438-461.
Langer, E. J. (1983). The psychology of control. Beverly Hills, Calif: Sage.
Larrick, R. P., and S. Blount (1995). Social context in tacit bargaining games. In R. M.
Kramer and D. M. Messick (Eds.), Negotiation as a social process, 268-284. Thou-
sand Oaks, Calif.: Sage.
Larrick, R. P., J. N. Morgan, and R. E. Nisbett (1990). Teaching the use of cost-benefit
reasoning in everyday life. Psychological Science, 1, 362-370.
Larrick, R. P., R. E. Nisbett, and J. N. Morgan (1993). Who uses the cost-benefit rules of
choice? Implications for the normative status of microcconomic theory. Organiza-
tional Behavior and Human Decision Processes, 56, 331-347.
Latane, B., and }. M. Darley (1970). The unresponsive bystander: Why doesn't he help?
New York: Appleton-Century-Crofts.
Latis, S. ). (1976). Method and appraisal in economics. New York: Cambridge University
Press.
Lawler, E. E., Ill (1981). Pay and organizational development. Reading, Mass.: Addison-
Wesley.
Lawler, E. E., Ill, S. A. Mohrman, and G. E. Lcdford, Jr. (1992). Employee involvement
and total quality management. San Francisco: Jossey-Bass.
Lawrence, P., and J. Lorsch (1967). Organization and environment. Boston: Harvard Uni-
versity, Graduate School of Business Administration, Division of Research.
Lax, D. A., and J. K. Sebenius (1986). The manager as negotiator. New York: Free Press.
Lazear, E. P. (1979). Why is there mandatory retirement? Journal of Political Economy,
87, 1261-1264.
Lazear, E. P. (1981). Agency, earnings profiles, productivity, and hours restrictions. Ameri-
can Economic Review, 71, 606-620.
Lazear, E. P. (1989). Pay equality and industrial politics. Journal of Political Economy, 97,
561-580.
Lazear, E. P. (1995). Personnel economics. Cambridge, Mass.: MIT Press.
Lazear, E. P., and S. Rosen (1981). Rank-order tournaments as optimum labor contracts.
Journal of Political Economy, 89, 841-864.
Lazersfeld, P. E, B. Berelson, and H. Camlet (1944). The people's choice. New York: Co-
lumbia University Press.
Ledford, G. E., Jr., E. E. Lawler III, and S. A. Mohrman (1995). Reward innovation
226 References
Louis Harris and Associates (1991). Laborforce 2000 survey. New York: Louis Harris and
Associates.
Lowin, A., and J. R. Craig (1968). The influence of level of performance on managerial
style: An experimental object lesson in the ambiguity of correlational data. Organi-
zational Behavior and Human Performance, 3, 440-458.
Luthans, R, and R. Kreitner (1975). Organizational behavior modification. Glenview, 111.:
Scott, Foresman.
Lykken, D. T., T. J. Bouchard, Jr., M. McGue, and A. Tellegen (1993). Heritability of
interests: A twin study. Journal of Applied Psychology, 78, 649-661.
MacDuffie, }. P. (1995). Human resource bundles and manufacturing performance: Flexi-
ble production systems in the world auto industry. Industrial and Labor Relations
Review, 48, 197-221.
MacKen/.ie, K. D., and R. House (1978). Paradigm development in the social sciences: A
proposed research strategy. Academy of Management Review, 3, 7-23.
Macy, B. A., and H. Izumi (1993). Organizational change, design, and work innovation: A
meta-arialysis of 131 North American field studies—1961-1991. Research in Orga-
nizational Change and Development: Vol. 7, 235-313. Greenwich, Conn.: JAI
Press.
Madden, J. F. (1985). The persistence of pay differentials: The economics of sex discrimi-
nation. In L. Larwood, A. II. Stromberg, and B. A. Gutek (Eds.), Women and Work:
An Annual Review. Vol. I, 76-114. Beverly Hills, Calif.: Sage.
Madison, D. L., R. W. Allen, L. W. Porter, P. A. Renwick, and B. T. Mayes (1980).
Organizational politics: An exploration of managers' perceptions. Human Relations,
33, 79-100.
Main, B. G. M., C. A. O'Reilly III, and J. Wade (1995). The CFO, the board of directors,
and executive compensation: Economic and psychological perspectives. Industrial
and Corporate Change, 4, 293-332.
Marine, H. G. (1965). Mergers and the market for corporate control. Journal of Political
Economy, 73, 110-120.
March, J. C., and J. G. March (1977). Almost random careers: The Wisconsin school
supcrintendcncy, 1940-1972. Administrative Science Quarterly, 22, 377-409.
March, J. G. (1966). The power of power. In D. Easton (Ed.), Varieties of political theory,
39-70. Englewood Cliffs, N.J.: Prentice Hall.
March, J. G. (1978). Bounded rationality, ambiguity, and the engineering of choice. Bell
Journal of Economics, 9, 587-608.
March, J. G., and J. P. Olsen (1984). The new institutionalism: Organizational factors in
political life. American Political Science Review, 78, 734-749.
March, J. G., and J. P. Olsen (1989). Rediscovering institutions: The organizational basis of
politics. New York: Free Press.
March, J. G., and H. A. Simon (1958). Organizations. New York: Wiley.
Marglin, S. A. (1974). What do bosses do? The origins and functions of hierarchy in
capitalist production. Review of Radical Political Economics, 6, 60-112.
Markus, II., and R. B. Zajonc (1985). The cognitive perspective in social psychology. In
G. Lind/,cy and E. Aronson (Eds.), Handbook of social psychology, 3rd ed.: Vol. 1,
Theory and method, 137-230. New York: Random House.
Marris, R. (1967). 'The economic theory of "managerial" capitalism. London: Macmillan.
Marsden, D., and R. Richardson (1994). Performing for pay? The effects of "merit pay"
on motivation in a public service. British Journal of Industrial Relations, 32,
219-242.
228 References
Martin, J. (1992a). Cultures in organizations: Three perspectives. New York: Oxford Univer-
sity Press.
Martin, J. (1992b). Eseaping the inherent conservatism of empirical organizational re-
search. In P. J. Frost and R. E. Stablcin (Eds.), Doing exemplary research, 233-239.
Ncwbury Park, Calif.: Sage.
Martin, J. (1994). The organization of exclusion: Institutionalization of sex inequality, gen-
dered faculty jobs, and gendered knowledge in organizational theory and research.
Organization, 1, 401-431.
Marwell, G. (1982). Altruism and the problem of collective action. In V. J. Derlega and J.
Gr/elak (Eds.), Cooperation and helping behavior: Theories and research, 207-226.
New York: Academic Press.
Marx, K. (trans. 1967). Capital, Vol. L New York: New World Paperbacks.
McCaffrey, D. P., S. R. Faerman, and D. W. Hart (1995). The appeal and difficulties of
participative systems. Organization Science, 6, 603-627.
McCain, B. R., C. A. O'Reilly, and J. Pfeffer (1983). The effects of departmental demogra-
phy on turnover: The case of a university. Academy of Management journal, 26,
626-641.
McGraw, K. O. (1978). The detrimental effects of reward on performance: A literature
review and a prediction model. In M. R. Lepper and D. Greene (Eds.), The hidden
costs of rewards: New perspectives on the psychology of human motivation. Hillsdale,
N.J.: Erlbaum.
McGregor, D. (1960). The human side of enterprise. New York: McGraw-Hill.
Medoff, J. L., and K. C. Abraham (1980). Experience, performance, and earnings. Quar-
terly journal of Economics, 95, 703-736.
Medoff, J. L., and K. C. Abraham (1981). Are those paid more really more productive?
The case of experience. Journal of Human Resources, 16, 186-216.
Mcindl, J. R., S. B. Ehrlich, and J. M. Dukerich (1985). The romance of leadership.
Administrative Science Quarterly, 30, 78-102.
Mellor, S. (1995). Gender composition and gender representation in local unions: Rela-
tionships between women's participation in local office and women's participation
in local activities. Journal of Applied Psychology, 80, 706-720.
Meyer, J. P., S. V. Paunonen, I. R. Gellatly, R. D. Coffin, and D. N. Jackson (1989).
Organizational commitment and job performance: It's the nature of the commit-
ment that counts, journal of Applied Psychology, 74, 152-1 56.
Meyer, ]. W., and B. Rowan (1977). Institutionalized organizations: Formal structure as
myth and ceremony. American Journal of Sociology, 83, 340-363.
Meyer, J. W., and W. R. Seott (1983). Organizational environments: Ritual and rationality.
Beverly Hills, Calif: Sage.
Meyer, ]. W., W. R. Scott, and T. E. Deal (1981). Institutional and technical sources of
organizational structure: Explaining the structure of educational organizations. In
II. D. Stein (Ed.), Organization and the human services, 151-178. Philadelphia:
Temple University Press.
Meyer, M. W. (1972a). Bureaucratic structure and authority. New York: Harper and Row.
Meyer, M. W. (1972b). Size and structure of organizations: A causal analysis. American
Sociological Review, 37, 434-440.
Meyer, M. W. (1975). Leadership and organizational structure. American Journal of Sociol-
ogy, 81, 514-542.
Meyer, M. W., and I,. G. Xncker (1989). Permanently failing organizations. Newbury Park,
Calif.: Sage.
References 229
Milgrom, P., and J. Roberts (1988). An economic approach to influence activities in orga-
nizations. American Journal of Sociology, 94, S154-S179.
Milgrorn, P., and J. Roberts (1992). Economics, organization, and management. Englewood
Cliffs, N.J.: Prentice Hall.
Miller, Daniel J. (1995). CEO salary increases may be rational after all: Referents and
contracts in CEO pay. Academy of Management journal, 38, 1361-1385.
Miner, J. B. (1984). The validity and usefulness of theories in an emerging organizational
science. Academy of Management Review, 9, 296-306.
Mint/berg, II. (1973). The nature of managerial work. New York: Harper and Row.
Mirvis, P. (1981). Personal correspondence: Review of Organizations and organization
theory.
Mischel, W. (1977). The interaction of person and situation. In D. Magnusson and N. S.
Endler (Eds.), Personality at the crossroads: Current issues in interactional psychol-
ogy, 333-352. Hillsdalc, N.J.: Erlbaum.
Mishel, L., and J. Bernstein (1995). The state of working America, J994-I995. Washington,
D.C.: Economic Policy Institute.
Mitchell, D. J. B., D. Lewin, and E. E. Lawler III (1990). Alternative pay systems, firm
performance, and productivity. In A. S. Blinder (Ed.), Paying for productivity: A
look at the evidence, 15-88. Washington, D.C.: Brooldngs.
Mitman, B. S. (1992). Theoretical and methodological issues in the study of organizational
demography and demographic change. In S. B. Bacharach and P. S. 'Iblbert (Eds.),
Research in the sociology of organizations: Vol. 10, 3-53. Greenwich, Conn.: JAI
Press.
Miyazawa, S. (1986). Legal departments of Japanese corporations in the United States: A
study on organizational adaptation to multiple environments. Kobe University Law
Review, 20, 97-162.
Mizruchi, M. S. (1992). The structure of corporate political action: Interfmn relationships
and their consequences. Cambridge, Mass.: Harvard University Press.
Mizruchi, M. S., and T. Koenig (1986). Economic sources of corporate political consen-
sus: An examination of interindustry relations. American Sociological Review, 51,
482-491.
Mizruchi, M. S., and L. B. Steams (1994). A longitudinal study of borrowing by large
American corporations. Administrative Science Quarterly, 39, 118-140.
Mohr, L. (1971). Organizational technology and organizational structure. Administrative
Science Quarterly, 16, 444-459.
Mohrman, S. A., S. G. Cohen, and A. M. Mohrman, Jr. (1995). Designing team-based
organizations: New forms for knowledge work. San Francisco: Jossey-Bass.
Molotch, H. L., and D. Boden (1985). Talking social structure: Discourse, domination,
and the Watergate hearings. American Sociological Review, 50, 273-288.
Mone, M. A., and W. McKinlcy (1993). The uniqueness value and its consequences for
organization studies. Journal of Management Inquiry, 2, 284-296.
Montemayor, E. F. (1995). Review of Alfie Kohn, Punished by Rewards: The Trouble with
Gold Stars, Incentives Plans, As, Praise, and Other Bribes. Personnel Psychology, 48,
941-945.
Moore, W. L., and J. Pfeffer (1980). The relationship between departmental power and
faculty careers on two campuses: The case for structural effect on faculty salaries.
Research in Higher Education, 13, 291-306.
Morgan, G. (Ed.) (1983). Beyond method. Beverly Hills, Calif.: Sage.
Morgan, G. (1990). Paradigm diversity in organizational research. In J. Hassard and D.
230 References
Pym (Eds.), The theory and philosophy of organizations: Critical issues and new
perspectives, 13-29. London: Routledge.
Morris, M. W., and K. Peng (1994). Culture and cause: American and Chinese attributions
for social and physical events. Journal of Personality and Social Psychology, 67, 949-
971.
Morris, S. (1973). Stalled professionalism: The recruitment of railway officials in the
United States, 1885-1940. Business History Review, 47, 317-334.
Mowday, R. T., !,. W. Porter, and R. M. Steers (1982). Employee-organizational linkages:
The psychology of commitment, absenteeism, and turnover. New York: Academic
Press.
Mowday, R. T., and D. G. Spencer (1981). The influence of task and personality charac-
teristics on employee turnover and absenteeism. Academy of Management Journal,
24, 634-642.
Mowday, R. T., and R. 1. Sutton (1993). Organizational behavior: Linking individuals and
groups to organizational contexts. Annual Review of Psychology, 44, 195-229.
Mueller, F. (1995). Organizational governance and employee cooperation: Can we learn
from economists? Human Relations, 48, 1217-1235.
Mumby, D. K. (1988). Communication and power in organizations: Discourse, ideology,
and domination. Norwood, N.J.: Ablex.
Murnby, D. K., and L. L. Putnam (1992). The politics of emotion: A feminist reading of
bounded rationality. Academy of Management Review, 17, 465-486.
Ncale, M. A. (1984). The effect of negotiation and arbitration cost salience on bargainer
behavior: The role of arbitrator and constituency in negotiator judgment. Organiza-
tional Behavior and Human Performance, 34, 97-111.
Nealc, M. A., and M. II. Bazerman (1985). The effects of framing and negotiator overcon-
fidcnce on bargainer behavior. Academy of Management Journal, 28, 34-49.
Neale, M. A., and M. H. Bazerman (1991). Cognition and rationality in negotiation. New
York: Free Press.
Neale, M. A., and G. B. Northcraft (1986). Experts, amateurs, and refrigerators: Compar-
ing expert and amateur decision making on a novel task. Organizational Behavior
and Human Decision Processes, 38, 305-317.
Neale, M. A., and G. B. Northcraft (1990). Experience, expertise, and decision bias in
negotiation: The role of strategic conceptualization. In B. II. Sheppard, M. H.
Bazerman, and R. J. Lewicki (Eds.), Research on negotiation in organizations: Vol.
2, 55—75. Greenwich, Conn.: JAI Press.
Newton, T., and T. Keenan (1991). Further analyses of the dispositional argument in
organizational behavior. Journal of Applied Psychology, 76, 781-787.
Nilakant, V., and H. Rao (1994). Agency theory and uncertainty in organizations: An eval-
uation. Organization Studies, 15, 649-672.
Nisbett, R. E., and L. Ross (1980). Human inference: Strategies and shortcomings of social
judgment. Englewood Cliffs, N.J.: Prentice Hall.
Nkomo, S. M. (1992). The emperor has no clothes: Rewriting "race in organizations."
Academy of Management Review, 17, 487-513.
Nord, W. (1969). Beyond the teaching machine: The neglected area of operant condition-
ing in the theory and practice of management. Organizational Behavior and Hu-
man Performance, 14, 371—397.
Northcraft, G. B., and M. A. Neale (1986). Opportunity costs and the framing of resource
allocation decisions. Organizational Behavior and Human Decision Processes, 37,
348-356.
Northcraft, G. B., and M. A. Ncalc (1987). Experts, amateurs, and real estate: An
References 231
O'Reilly, C. A., R. C. Snycler, and J. N. Boothe (1993). Effects of executive team demogra-
phy on organizational change. In G. P. Huber and W. H. Click (Eds.), Organiza-
tional change and redesign, 147-175. New York: Oxford University Press.
O'Reilly, C. A., and B. A. Weitz (1980). Managing marginal employees: The use of warn-
ings and dismissals. Administrative Science Quarterly, 25, 467-484.
Osterman, P. (1994). How common is workplace transformation and who adopts it? Indus-
trial and Labor Relations Review, 47, 173-188.
Palmer, D. P. (1983). Broken tics: Interlocking directorates and intercorporate coordina-
tion. Administrative Science Quarterly, 28, 40—55.
Palmer, D. P., B. M. Barber, X. Zhou, and Y. Soysal (1995). The friendly and predatory
acquisitions of large U.S. corporations in the 1960s: The other contested terrain.
American Sociological Review, 60, 469-499.
Palmer, D. P., R. Ericdland, and J. Singh (1986). The ties that bind: Organizational and
class bases of stability in a corporate interlock network. American Sociological Re-
view, 51, 781-796.
Palmer, D. P., P. D. Jennings, and X. Zhou (1993). Late adoption of the multidivisional
form by large U.S. corporations: Institutional, political, and economics accounts.
Administrative Science Quarterly, 38, 100-131.
Parker, M., and J. Slaughter (1988). Management by stress. Technology Review, 91, 38-44.
Parker, R. (1993). Can economists save economics? American Prospect, 13, 148-160.
Pascale, R. (1985). The paradox of "corporate culture": Reconciling ourselves to socializa-
tion. California Management Review, 27, 26-41.
Pascale, R. (1990). Managing on the edge. New York: Simon and Schuster.
Pearcc, J. L, W. B. Stevenson, and J. I,. Perry (1985). Managerial compensation based on
organizational performance: A time scries analysis of the effects of merit pay. Acad-
emy of Management Journal, 28, 261-278.
Pennings, J. (1975). The relevance of the structural-contingency model of organizational
effectiveness. Administrative Science Quarterly, 30, 393-410.
Perrow, C. (1970a). Departmental power and perspectives in industrial firms. In M. N.
Zakl (Ed.), Power in organizations, 59-89. Nashville: Vanderbilt University Press.
Perrow, C. (1970b). Organizational analysis: A sociological view. Belmont, Calif.: Wads-
worth.
Perrow, C. (1992). Organisational theorists in a society of organisations. International Soci-
ology, 7, 371-379.
Pervin, L. A. (1985). Personality: Current controversies, issues, and directions. Annual re-
view of psychology, 36, 83-114.
Pervin, L. A., and M. Lewis (1978). Overview of the internal-external issue. In L,. A. Pervin
and M. Lewis (Eds.), Perspectives in interactional psychology, 1—22. New York:
Plenum.
Peters, T. J. (1978). Symbols, patterns, and settings: An optimistic case for getting things
done. Organizational Dynamics, 7, 3-23.
Pettigrcw, A. M. (1972). Information control as a power resource. Sociology, 6, 187—204.
Pettigrew, A. M. (1973). Politics of organizational decision-making. London: Tavistock.
Pcttigrew, A. M. (1979). On studying organizational cultures. Administrative Science Quar-
terly, 24, 570-581.
Pfeffer, J. (1972a). Interorganizational influence and managerial attitudes. Academy of
Management Journal, 15, 317-330.
Pfeffer, J. (1972b). Merger as a response to organizational interdependence. Administrative
Science Quarterly, 17, 382-394.
Pfeffer, J. (1972c). Size and composition of corporate boards of directors: The organization
and its environment. Administrative Science Quarterly, 17, 218-228.
References 233
Pfeffer, J. (1973). Size, composition, and function of hospital boards of directors: A study of
organization-environment linkage. Administrative Science Quarterly, 18, 349-364.
Pfeffer, J. (1977). The ambiguity of leadership. Academy of Management Review, 2, 104-
112.
Pfeffer, J. (1981a). Four laws of organizational research. In A. Van dc Ven and W. Joyce
(Eds.), Perspectives on organization design and behavior, 409-418. New York: Wiley.
Pfeffer, J. (1981b). Management as symbolic action: The creation and maintenance of
organizational paradigms. In L. L. Cummings and B. M. Staw (cds.), Research in
Organizational Behavior: Vol. 3, 1—52. Greenwich, Conn.: JAI Press.
Pfeffer, J. (1981c). Power in organizations. Marshfickl, Mass.: Pitman.
Pfeffer, J. (1982). Organizations and organization theory. Marshfield, Mass.: Pitman.
Pfeffer, J. (1983). Organizational demography. In L. L. Cummings and B. M. Staw (Eds.),
Research in organizational behavior: Vol. 5, 299-357. Greenwich, Conn.: JAI Press.
Pfeffer, J. (1985). Organizations and organization theory. In G. Lindzey and E. Aronson,
The handbook of social psychology, 3rd ed., 379-440. New York: McGraw-Hill.
Pfeffer, J. (1987). A resource dependence perspective on intercorporate relations. In M. S.
Mizruchi and M. Schwartz (Eds.), Intercorporate relations: The structural analysis
of business, 25-55. Cambridge: Cambridge University Press.
Pfeffer, J. (1991). Organization theory and structural perspectives on management. Journal
of Management, 17, 789-803.
Pfeffer, J. (1992). Managing with power: Politics and influence in organizations. Boston:
Harvard Business School Press.
Pfeffer, J. (1993). Barriers to the advance of organizational science: Paradigm development
as a dependent variable. Academy of Management Review, 18, 599-620.
Pfeffer, J. (1994). Competitive advantage through people: Unleashing the power of the work
force. Boston: Harvard Business School Press.
Pfeffer, J., and J. N. Baron (1988). Taking the workers back out: Recent trends in the
structuring of employment. Research in organizational behavior: Vol. 10, 257-303.
Greenwich, Conn.: JAI Press.
Pfeffer, J., and Y. Cohen (1984). Determinants of internal labor markets in organizations.
Administrative Science Quarterly, 29, 550-572.
Pfeffer, J., and A. Davis-Blake (1986). Administrative succession and organizational perfor-
mance: How administrator experience mediates the succession effect. Academy of
Management Journal, 29, 72-83.
Pfeffer, J., and A. Davis-Blake (1987). The effect of the proportion of women on salaries:
The case of college administrators. Administrative Science Quarterly, 32, 1-24.
Pfeffer, J., and A. Davis-Blake (1990). Determinants of salary dispersion in organizations.
Industrial Relations, 29, 38-57.
Pfeffer, J., and A. Davis-Blake (1992). Salary dispersion, location in the salary distribution,
and turnover among college administrators. Industrial and Labor Relations Review,
45, 753-763.
Pfeffer, J., and A. M. Konrad (1991). The effects of individual power on earnings. Work
and Occupations, 18, 385-414.
Pfeffer, J., and N. Langton (1988). Wage inequality and the organization of work: The case
of academic departments. Administrative Science Quarterly, 33, 588-606.
Pfeffer, J., and N. Langton (1993). The effect of wage dispersion on satisfaction, productiv-
ity, and working collaboratively: Evidence from college and university faculty. Ad-
ministrative Science Quarterly, 38, 382—407.
Pfeffer, J., and J. Lawler (1980). Effects of job alternatives, extrinsic rewards, and behavioral
commitment on attitude toward the organization: A field test of the insufficient
justification paradigm. Administrative Science Quarterly, 25, 38-56.
234 References
Pfeffer, J., and H. Leblebici (1973). The effect of competition on some dimensions of
organizational structure. Social Forces, 52, 268-279.
Pfeffer, J., A. Leong, and K. Strehl (1977). Paradigm, development, and particularism:
Journal publication in three scientific disciplines. Social Forces, 55, 938-951.
Pfeffer, J., and W. L. Moore (1980a). Average tenure of academic department heads: The
effects of paradigm, si/.c, and departmental demography. Administrative Science
Quarterly, 25, 387-406.
Pfeffer, J., and W. L. Moore (1980b). Power in university budgeting: A replication and
extension. Administrative Science Quarterly, 25, 637-653.
Pfeffer, J., and P. Nowak (1976). Joint ventures and interorganizational interdependence.
Administrative Science Quarterly, 21, 398-418.
Pfeffer, J., and C. A. O'Reilly (1987). Hospital demography and turnover among nurses.
Industrial Relations, 26, 158-173.
Pfeffer, J., and J. Ross (1982). The effects of marriage and a working wife on occupational
and wage attainment. Administrative Science Quarterly, 27, 66-80.
Pfeffer, J., and G. R. Salancik (1974). Organizational decision making as a political pro-
cess: The case of a university budget. Administrative Science Quarterly, 19, 135-
151.
Pfeffer, J., and G. R. Salaneik (1977). Administrator effectiveness: The effects of advocacy
and information on resource allocations. Human Relations, 30, 641-656.
Pfeffer, J., and G. R. Salancik (1978). The external control of organizations: A resource
dependence perspective. New York: Harper and Row.
Pfeffer, J., G. R. Salancik, and H. Leblebici (1976). The effect of uncertainty on the
use of social influence in organizational decision making. Administrative Science
Quarterly, 21, 227-245.
Pil, F. K., and J. P. MacDnffie (1996). The adoption of high-involvement work practices.
Industrial Relations, 35, 423-455.
Podolny, J. M. (1993). A status-based model of market competition. American journal of
Sociology, 98, 829-872.
Podolny, J. M. (1994). Market uncertainty and the social character of economic exchange.
Administrative Science Quarterly, 39, 458-483.
Podolny, J. M., and J. N. Baron (1995, August). Resources and relationships: Social net-
works, mobility, and satisfaction in the work place. Paper presented at the Annual
Meeting of the Academy of Management, Vancouver, Ganacla.
Pondy, L. R. (1978). Leadership is a language game. In M. W. McGall, Jr., and M. M.
Lombardo (Eds.), Leadership: Where else can we go? 87-99. Durham, N.C.: Duke
University Press.
Porac, J. R, and G. R. Salancik (1981). Generic ovcrjustification: The interaction of extrin-
sic rewards. Organizational Behavior and Human Performance, 27, 197-212.
Preslhus, R. (1978). The organizational society, rev. ed. New York: St. Martin's Press.
Pugh, D. S., D. J. Iliekson, G. R. Hinings, and C. Turner (1968). Dimensions of organiza-
tion structure. Administrative Science Quarterly, 13, 65-105.
Pugh, D. S., D. J. Iliekson, C. R. Minings, and C. Turner (1969). The context of organiza-
tion structures. Administrative Science Quarterly, 14, 91—114.
Rafach, A., and R. 1. Sutton (1991). Emotional contrast strategies as means of social influ-
ence: Lessons from criminal interrogators and bill collectors. Academy of Manage-
ment Journal, 34, 749-775.
Rankin, T. (1986). Integrating QWL and collective bargaining. Work-Life Review, 5, 14-
18.
Rcc, M. J., and J. A. Earles (1991). Predicting training success: Not much more than g.
Personnel Psychology, 44, 321-332.
References 235
Schneider, B. (1987). The people make the place. Personnel Psychology, 40, 437-453.
Schneider, B., II. W. Goldstein, and D. B. Smith (1995). The ASA framework: An update.
Personnel Psychology, 48, 747-773.
Schoonhovcn, C. B. (1981). Problems with contingency theory: Testing assumptions hid-
den within the language of contingency "theory." Administrative Science Quarterly,
26, 349-377.
Schoorman, F. D. (1988). Escalation bias in performance appraisals: An unintended conse-
quence of supervisor participation in hiring decisions. Journal of Applied Psychol-
ogy, 73, 58-62.
Schwartz, B. (1974). Waiting, exchange, and power: The distribution of time in social
systems. American journal of Sociology, 79, 841-870.
Scott, W. R. (1987). The adolescence of institutional theory. Administrative Science Quar-
terly, 32,493-511.
Scott, W. R. (1992). Organizations: Rational, natural, and open systems, 3rd ed. Englewood
Cliffs, N.J.: Prentice Hall.
Scott, W. R. (1995). Institutions and organizations. Thousand Oaks, Calif.: Sage.
Scully, C. W. (1974). Discrimination: The case of baseball. In R. G. Noll (Ed.), Govern-
ment and the sports business, 221-273. Washington, D.C.: Brookings.
Seaman, R. (1995). How self-directed work teams support strategic alignment. Compensa-
tion and Benefits Review, 27, 23-32.
Sears, D. O., and C. L. Funk (1990). Self-interest in Americans' political opinions. In J. ].
Mansbridge (Ed.), Beyond self-interest, 147-170. Chicago: University of Chicago
Press.
Scars, D. O., R. R. Lau, T. R. Tyler, and H. M. Allen, Jr. (1980). Self-interest vs. symbolic
politics in policy attitudes and presidential voting. American Political Science Re-
view, 74, 670-684.
Seligman, D. L. (1992). A question of intelligence: The IQ debate in America. New York:
Birch Lane Press.
Selznick, P. (1949). TVA and the grass roots. Berkeley: University of California Press.
Selznick, P. (1957). leadership in administration. Evanston, 111.: Row, Peterson.
Sen, A. K. (1977). Rational fools. Philosophy and Public Affairs, 6, 317-344.
Shadur, M. A., J. J. Rodwcll, and G. j. Bamber (1995). Factors predicting employees'
approval of lean production. Human Relations, 48, 1403-1425.
Shamir, B., R. J. House, and M. Arthur (1993). The motivational effects of charismatic
leadership: A self-concept based theory. Organization Science, 4, 577-594.
Sheldon, M. E. (1971). Investments and involvements as mechanisms producing commit-
ment in organizations. Administrative Science Quarterly, 16, 143-150.
Shell, R. I,., and R. G. Allgeier (1992). A multi-level incentive model for service organiza-
tions. Applied Ergonomics, 23, 49-53.
Shenhav, Y. (1995). From chaos to systems: The engineering foundations of organization
theory, 1879-1932. Administrative Science Quarterly, 40, 557-585.
Shenhav, Y., and Y. Ilaberfeld (1992). Paradigm uncertainty, gender composition, and
earnings inequality in scientific disciplines: A longitudinal study, 1972—1982. In P.
Tolbert and S. B. Bacharach (Eds.), Research in the sociology of organizations: Vol.
10, 141-172. Greenwich, Conn.: JAI Press.
Sheriff, P., and E. J. Campbell (1992). Room for women: A case study in the sociology
of organizations. In A. J. Mills and P. Tancrcd (Eds.), Gendering Organizational
Analysis, 31-45. Newbury Park, Calif: Sage.
Shore, L. M., K. Barksdale, and T. H. Shore (1995). Managerial perceptions of employee
commitment to the organization. Academy of Management Journal, 38, 1593-1615.
References 237
Shore, L. M., and S. J. Wayne (1993). Commitment and employee behavior: Comparison
of affective commitment and continuance commitment with perceived organiza-
tional support. Journal of Applied Psychology, 78, 774-780.
Sichl, C., and J. Martin (1990). Organizational culture: A key to financial performance?
In B. Schneider (Ed.), Organizational climate and culture, 241-281. San Francisco:
Jossey-Bass.
Sigelman, L., H. Milward, and J. M. Shepard (1982). The salary differential between male
and female administrators: Equal pay for equal work? Academy of Management
Journal, 25, 664-671.
Sills, D. L. (1957). The volunteers. Glencoe, 111.: Free Press.
Simon, II. A. (1947). Administrative behavior. New York: Macmillan.
Simon, II. A. (1963). Discussion: Problems of methodology. American Economic Review
Proceedings, 53, 229-231.
Simon, H. A. (1964). On the concept of organizational goal. Administrative Science Quar-
terly, 9, 1-22.
Simon, H. A. (1978). Rationality as process and as product of thought. American Economic
Review, 68, 1-16.
Simon, II. A. (1979). Rational decision making business organizations. American Economic
Review, 69, 493-513.
Simon, H. A. (1985). Human nature in politics: The dialogue of psychology with political
science. American Political Science Review, 79, 293-304.
Simon, II. A. (1991). Organizations and markets. Journal of Economic Perspectives, 5, 25-
44.
Sims, H., A. Szilagyi, and R. Keller (1976). The measurement of job characteristics. Acad-
emy of Management Journal, 19, 195-212.
Singh, J. V., and C. J. Lumsden (1990). Theory and research in organizational ecology.
Annual Review of Sociology, 16, 161-195.
Skinner, B. F. (1953). Science and human behavior. New York: Macmillan.
Smircich, L. (1983). Concepts of culture and organizational analysis. Administrative Sci-
ence Quarterly, 28, 339-359.
Smith, K. G., K. A. Smith, }. D. Olian, H. P. Sims, jr., D. P. O'Bannon, and J. A.
Scully (1994). Top management team demography and process: The role of social
integration and communication. Administrative Science Quarterly, 39, 412-
438.
Smith, M. J., P. Carayon, K. J. Sanders, S. Y. Lim, and D. LeGrande (1992). Employee
stress and health complaints in jobs with and without electronic performance moni-
toring. Applied Ergonomics, 23, 17-28.
Soref, M., and M. Zeitlin (1987). Finance capital and the internal structure of the capital-
ist class in the United States. In M. Mizruchi and M. Schwartz (Eds.), Intercorpo-
rate relations: The structural analysis of business, 56-84. Cambridge: Cambridge
University Press.
South, S. J., C. M. Bonjean, W. T. Markham, and J. Corder (1982). Social structure
and intergroup interaction: Men and women of the federal bureaucracy. American
Sociological Review, 47, 587-599.
Spangler, E., M. A. Gordon, and R. M. Pipkin (1978). Token women: An empirical test
of Kanter's hypothesis. American Journal of Sociology, 84, 160-170.
Spangler, W. D., and R. J. House (1991). Presidential effectiveness and the leadership
motive profile. Journal of Personality and Social Psychology, 60, 439-455.
Spence, A. M. (1974). Market signaling: Informational transfer in hiring and related pro-
cesses. Cambridge, Mass.: Harvard University Press.
238 References
Westphal, J. D., and E. J. Zajac (1995). Who shall govern? CEO/board power, demo-
graphic similarity, and new director selection. Administrative Science Quarterly, 40,
60-83.
Wharton, A. S. (1992). The social construction of gender and race in organizations: A
social identity and group mobilization perspective. In P. Tolbert and S. B. Bachar-
ach (Eds.), Research in the sociology of organizations: Vol. 10, 55-84. Greenwich,
Conn.: JAI Press.
Wharton, A. S., and J. N. Baron (1987). So happy together? The impact of gender segrega-
tion on men at work. American Sociological Review, 52, 574-587.
White, S., and T. Mitchell (1979). Job enrichment versus social cues: A comparison and
competitive test. Journal of Applied Psychology, 64, 1-9.
Whitt, J. A. (1979). Toward a class-dialectical model of power: An empirical assessment of
three competing models of political power. American Sociological Review, 44, 81-
100.
Whitt, J. A. (1980). Can capitalists organize themselves? In G. W. Domhoff (Ed.), Power
structure research, 97—113. Beverly Hills, Calif.: Sage.
Wholey, D. R., J. B. Christiansen, and S. M. Sanchez (1992). Organization size and failure
among health maintenance organizations. American Sociological Review, 57, 829-
842.
Wholey, D. R., J. B. Christiansen, and S. M. Sanchez (1993). The effects of physician and
corporate interests on the formation of health maintenance organizations. American
Journal of Sociology, 99, 164-200.
Wiersema, M. E, and K. A. Bantel (1992). Top management team demography and corpo-
rate strategic change. Academy of Management Journal, 35, 91-121.
Wiersema, M. E, and A. Bird (1993). Organizational demography in Japanese firms:
Group heterogeneity, individual dissimilarity, and top management team turnover.
Academy of Management Journal, 36, 996-1025.
Williamson, O. E. (1964). Corporate control and business behavior: Managerial objectives
in a theory of the firm. Englewood Cliffs, N.J.: Prentice Hall.
Williamson, O. E. (1975). Markets and hierarchies: Analysis and antitrust implications.
New York: Free Press.
Williamson, O. E. (1985). The economic institutions of capitalism. New York: Free Press.
Williamson, O. E. (1991). Economic institutions: Spontaneous and intentional gover-
nance. Journal of Law, Economics, and Organization, 7, 159-187.
Williamson, O. E. (1993). Opportunism and its critics. Managerial and Decision Econom-
ics, 14, 97-107.
Williamson, O. E. (1995). Hierarchies, markets, and power in the economy: An economic
perspective. Industrial and Corporate Change, 4, 21-49.
Williamson, O. I1',., and W. G. Ouchi (1981). The markets and hierarchies program of
research: Origins, implications, prospects. In A. II. Van de Ven and W. F. Joyce
(Eds.), Perspectives on organizational design and behavior, 347-370. New York:
Wiley.
Wilson, T. D., and G. D. Lassiter (1982). Increasing intrinsic interest with superfluous
external constraints, journal of Personality and Social Psychology, 42, 811-819.
Winter, D. G. (1987). Leader appeal, leader performance, and the motive profiles of lead-
ers and followers: A study of American presidents and elections. Journal of Personal-
ity and Social Psychology, 52, 196-202.
Wolf, W. C., and N. D. Fligstcin (1979). Sex and authority in the workplace: The causes
of sexual inequality. American Sociological Review, 44, 235—252.
References 243
Womack, J. P., D. T. Jones, and D. Roos (1990). The machine that changed the world.
New York: Rawson Associates.
Wood, S. (1996). High commitment management and payment systems. ]oumal of Man-
agement Studies, 33, 53-77.
Woodward, J. (1965). Industrial organization: Theory and practice. London: Oxford Univer-
sity Press.
Woodward, J. (Ed.) (1970). Industrial organization: Behaviour and control. London: Oxford
University Press.
Woolhandler, S., and D. U. Himmelstein (1991). The deteriorating administrative effi-
ciency of the U.S. health care system. New England journal of Medicine, 324,
1253-1258.
Wright, E. O., and L. Perrone (1977). Marxist class categories and income inequality.
American Sociological Review, 42, 32-55.
Wrong, D. H. (1968). Some problems in defining social power. American Journal of Sociol-
ogy, 73, 673-681.
Wyatt Company (1993). Wyatt's 1993 survey of corporate restructuring-best practices in cor-
porate restructuring. New York: Wyatt Company.
Yoels, W. C. (1974). The structure of scientific fields and the allocation of editorships on
scientific journals: Some observations on the politics of knowledge. Sociological
Quarterly, 15, 264-276.
Zajonc, R. B. (1980). Cognition and social cognition: A historical perspective. In L. Fes-
tinger (Ed.), Four decades of social psychology, 180-204. New York: Oxford Univer-
sity Press.
Zald, M. N. (1993). Toward a reconceptualization of the foundations of the field. Organi-
zation Science, 4, 513-528.
Zammuto, R. F., and T. Connolly (1984). Coping with disciplinary fragmentation. Organi-
zational Behavior Teaching Review, 9, 30-37.
Zeitlin, M. (1974). Corporate ownership and control: The large corporation and the capi-
talist class. American Journal of Sociology, 79, 1073-1119.
Zenger, T. R. (1992). Why do employers only reward extreme performance? Examining
the relationship among performance, pay, and turnover. Administrative Science
Quarterly, 37, 198-219.
Zenger, T. R., and B. S. Lawrence (1989). Organizational demography: The differential
effects of age and tenure distributions on technical communication. Academy of
Management Journal, 32, 353-376.
Zucker, L. G. (1983). Organizations as institutions. In S. B. Bacharach (Eel.), Research in
the sociology of organizations: Vol. 2, 1-47. Greenwich, Conn.: JAI Press.
Zucker, L. G. (1989). Combining institutional theory and population ecology: No legiti-
macy, no history. American Sociological Review, 54, 542-545.
This page intentionally left blank
INDEX
autonomy. See also rewards and incentives bundle system, defined, 174
absenteeism and, 38 business schools
achievement and, 38 in Europe, 13
dispositions and, 37 organization studies in, 9, 11, 13, 14-17,
management and, 21-23, 37 18, 189-90
pension funds and, 22 relevance and, 13
size distribution and, in United States, 13
autonomy-control, denned, 182 bystander style, defined, 148. See also
power and influence
bankruptcy, 59
barriers, 199-200 capital markets. See externalization; inter-
behavior nalization
causality of, 25-41 career chances. See mobility
in China, 26-27 career mobility. See mobility
cognitive model of, 77—79 careers. See job attitudes; job choice; job
commitment and, 116 complexity; job design; job interests;
control of, 100-135 job performance; job satisfaction; job
demography and, 98 security; job success
design and, 25, 26 causality
economic models of, 44-55 of behavior, 25-41
embedded nature of, 55 high-commitment work system and, 171
individual differences and, 27-33 leadership traits and, 28
information and, 111 locus of, 5
interaction and, 55-56 normative order and, 26-27
learning perspective and, 25 organization studies and, 5
models of, 5, 42-80 cause map
moral model of, 73-77 defined, 78
normative order and, 26—27 origins of, 79
operant-conditioning perspective and, 25 centrality, 145
organization studies and, 13-14 CKOs. See also management
rationality of, 44 background in finance of, 23
retrospectively rational model of, 65-73 boards of directors and, 109
situationalist perspective of, 25 earnings of, 23, 109-10
social model of, 5, 55-65 in France, 23
structure and, 81 in Germany, 23
surveillance of, 114-16 in Japan, 23
in United States, 26-27 performance and, 109, 114
behavioral commitment. See commitment rewards and incentives and, 114
belief systems, 175 size of organizations and, 109
Big Five. See also dispositions; emotions; tournament model and, 110
tenure in United Kingdom, 23
criticism of, 34 in United States, 23
types of, 32 challenges, 37-38
birth. See founding and mortality chance, 138-39
blacks. See minorities; race change. See innovation and change
boards of directors, 109. See also manage- characteristics. See dispositions
ment charismatic leaders. See transformational
boundaries leadership
heightening of, 83 chief executive officers. See CEOs
permeability of, 9 China, behavior in, 26-27
Index 247
symbolic predispositions and, 74-75 smallness, liability in, 165, 168. See also
self-justification model. See also retrospec- size of organizations
tively rational model social behavior, defined, 79. See also be-
commitment and, 67 havior
economies and, 70 social capital, defined, 56, 98
self-interest model and, 74 social class, 184-86
self-perception model. See also retrospec- social cognition. See cognitive model
tively rational model social composition
cognitive dissonance and, 66 effects of, 81
commitment and, 66, 116 of organizations, 81-99
criticism of, 67 structure and, 81
economics and, 67 social contagion, 60-61
incentives and, 66-67 social control. See control
premises of, 66 social identity, 153
surveillance and, 70 social influences, 5-6
self-selection, 36. See also selection social integration, 84-85
senior corporate executives, 23. See also socialization
management commitment and, 118
seniority. See tenure defined, 118
seniority-based wage systems, 49. See also effects of, 120
earnings goals of, 118-19
sent interlocks, defined, 64. See also net- leadership and, 127
works process of, 118-20
"servants of power," 177-78 tactics of, 120
service organizations, 21. See also organiza- social model. See also institutional theory;
tions resource dependence theory
sex. See gender of behavior, 55-65
sexual harassment, 94. See also gender characteristics of, 176
shotgun style, defined, 148. See also power cognitive model and, 77
and influence criticism of, 64-65
signals, defined, 101 defined, 55
similarity. See heterogeneity demography and, 98-99
Simon, Herbert, 12-13 economic model compared to, 64
situational effects, 30-31 imitation and, 56-58
situationalist perspective interaction and, 55-56
of behavior, 25 matrix management and, 57—58
defined, 25 negotiation and, 153
job satisfaction and, 29 networks and, 56-58, 60
legal and political environment and, 25 opportunities and, 82
relative variance and, 28-29 organizations as units and, 56-57
situations. See design personal contacts and, 56
size distribution, 21 poison pills and, 57
size of organizations. See also liability in power and influence and, 82
smallness structure and, 81
competition and, 167 turnover and, 56
differentiation and, 161 social policy, 41
resource partitioning and, 168 social psychology. See also psychology
variation and, 161 characteristics of, 12
skill, 182. See also deskilling normative control and, 102
262 Index